Effects of Environmental Contaminants on Gene Expression, DNA Methylation and Gut Microbiota in Buff-Tailed Bumble Bee - Bombus Terrestris

UNIVERSITYOF LEICESTER

DOCTORAL THESIS

Effects of environmental contaminants on gene expression, DNA methylation and gut microbiota in Buff-tailed Bumble bee - Bombus terrestris

Author: Supervisor:

Pshtiwan BEBANE Eamonn MALLON

A thesis submitted in fulﬁlment of the requirements for the degree of Doctor of Philosophy

in the

Department of Genetics and Genome Biology

March 2019 Abstract

Bee populations are increasingly at risk. In this thesis, I explore various mechanisms through which environmental contaminants, namely imidacloprid and black carbon, can affect bumble bee epigenetics, behaviour and gut microbiota.

I found imidacloprid has numerous epigenetic effects on Bombus terrestris non reproductive workers. I analysed three whole methylome (BS-seq) libraries and seven RNA-seq libraries of the brains of imidacloprid exposed workers and three BS-seq libraries and nine RNA-seq libraries from unexposed, control workers. I found 79, 86 and 16 genes differentially methylated at CpGs, CHHs and CHGs sites respectively between groups. I found CpG methylation much more focused in exon regions compared with methylation at CHH or CHG sites. I found 378 genes that were differentially expressed between imidacloprid treated and control bees. In addition, I found 25 genes differentially alternatively spliced between control and imidacloprid samples.

I used Drosophila melanogaster as a model for the behavioural effects of imidacloprid on insects. Imidacloprid did not affect ﬂies’ periodicity. Low doses (2.5 ppb) of imidacloprid increased ﬂies’ activity while high doses (20 ppb) decreased activity. Canton-S strain was more sensitive to imidacloprid during geotaxis assay than M1217.

I proposed that a possible modulator of imidacloprid’s effects on insects is its effects on insects’ gut microbiota. I produced and analysed thirty libraries of 16S rRNA metagenomic data from bee guts dividing by control and imidacloprid. There were no differences in alpha and beta bacterial diversity between groups while the ratio of one species of latobacillus increased in bees exposed to imidacloprid.

As an extension of this microbiota work, I researched the effect of black carbon on gut microbiota in the bumble bee. Bacterial cultivation methods showed significant increases of species (CFU) in bees exposed to black carbon and this was also detected by a qPCR approach. How- ever, in a metagenomic analysis, I did not find significant differences in alpha and beta diversity richness between black carbon and control groups. Declaration of Authorship

I, Pshtiwan BEBANE, declare that this thesis titled, ’Effects of environmental contaminants on gene expression, DNA methylation and gut microbiota in buff-tailed bumble bee Bombus terrestris’ and the work presented in it are my own. I conﬁrm that:

■ This work was done wholly or mainly while in candidature for a research degree at this University.

■ Where any part of this thesis has previously been submitted for a degree or any other qualiﬁcation at this University or any other institution, this has been clearly stated.

■ Where I have consulted the published work of others, this is always clearly attributed.

■ Where I have quoted from the work of others, the source is always given. With the exception of such quotations, this thesis is entirely my own work.

■ I have acknowledged all main sources of help.

■ Where the thesis is based on work done by myself jointly with others, I have made clear exactly what was done by others and what I have contributed myself.

Signed:

Date:

ii DEDICATION

This thesis is dedicated to:

• My beloved, dearest and brilliant wife “Jwan” who never got tired of encourag- ing me and who leads me through the valley of darkness with light of hope and support.

• My lovely little guys, my sons “Yousif” and “Laween”, whom I can’t force myself to stop loving.

• My great parents, who never stop giving of themselves in countless ways.

• My beloved brothers (shaxawan, Baxwan, Sherwan, Nahrow) and sisters (Amena and Shermen).

• My great Parents-in-law.

• My beloved brothers-in-law (Aso, Karzan, Karwan) and sisters-in-law (Hero, Shena and Shema).

• My friends who encourage and support me.

• All of the people in my life who have touched my heart. Acknowledgements

Alhamdulillah, thank you Allah for all the blessing, You had given me. Thank You again for giving me the opportunity, knowledge and courage to overcome challenges and ﬁn- ish this PhD project.

I would like to thank and express my sincere gratitude to my supervisor Prof Eamonn Mallon. Thank you for all your support and guidance throughout my PhD, for your pa- tience, motivation, and immense knowledge. I would like to thank my second supervisor Prof Ezio Rosato who kindly continuous encouragement and has a nice suggestion during my study.

Besides my supervisors, I would like to thank my PRP members Dr Sinéad Drea and Dr Iain Barber for their intuitive comments, guidance, questions and also suggestion.

I would like to express my deepest thanks to every member of our lab past and present, with particular nods to Dr Mirko Pegoraro, Dr.Benjamin Hunt, Alun Jones, Hollie Mar- shall, Boris Berkhout, Christian Thomas, Zoe Lonsdale, all of whom were generous with time, assistance and opinions. Special thanks to the lab technicians, particularly Lisa Bedder and all staff, students in lab 104 and 121.

Also, I want to thank all people in Genetics and Genome Biology, lab mates and admin- istrative staff for their continuous support. Very special thanks also to the University of Leicester for giving me an opportunity to complete this study.

I would like to thank my parents, father-in-law, brothers, relatives, Ministry of Higher Education in Kurdistan/HCDP and Koya University for sponsoring and ﬁnancially sup- porting me throughout my PhD study.

I would especially like to thank my family. My dearest wife, Jwan has been extremely supportive of me throughout this entire process and has made countless sacriﬁces to help me get to this point. My mum, dad, sisters and brothers deserve special thanks for their continued support and encouragement. Without such a team behind me, I doubt that I would be in this place today.

i Contents

Abstract i

Declaration of Authorship ii

Acknowledgementsi

List of Figures vi

List of Tables xiii

Abbreviations xv

1 General introduction1 1.1 Interactions between plants-pollinators-pests-ecotoxicology...... 1 1.2 Bumble bees and their importance in nature...... 3 1.3 Evidence for the decline in bee populations...... 5 1.4 Factors affecting insects communities...... 6 1.4.1 Black carbon...... 6 1.4.2 What are neonicotinoids?...... 7 1.4.2.1 Imidacloprid and environmental residues...... 9 1.4.2.2 Lethal effect of neonicotinoids...... 10 1.4.2.3 Sub-lethal effects of neonicotinoids...... 11 Non neural effects...... 12 Gut bacteria effect...... 13 1.4.3 Epigenetic effects...... 14 1.5 Aim and justiﬁcation...... 15

2 The neonicotinoid, imidicloprid affects gene expression, alternative splicing and DNA methylation in Bombus terrestris 17 2.1 Introduction...... 18 2.2 Materials and methods...... 21 2.2.1 Bee husbandry...... 21

ii Contents iii

2.2.2 Neonicotinoid feeding and brain sampling...... 21 2.2.3 BS-seq...... 21 2.2.3.1 Genomic DNA extraction, sequencing and mapping.... 21 2.2.3.2 Methylation differences between treatments...... 22 2.2.3.3 Methylation differences in exon level...... 23 2.2.4 RNA-seq...... 23 2.2.4.1 RNA extraction and Illumina sequencing...... 23 2.2.4.2 Alignment and assemble transcripts...... 24 2.2.4.3 Differential gene expression analysis...... 24 2.2.4.4 Alternative splicing events...... 24 2.2.5 GO term enrichment and Kegg analysis...... 25 2.3 Results...... 26 2.3.1 Methylation...... 26 2.3.1.1 Sequencing, alignment and Methylation analysis...... 26 2.3.1.2 Methylation differences between treatments...... 28 2.3.2 Methylation differences in exon level...... 31 2.3.3 GO term and KEGG pathway analysis...... 34 2.3.4 Differential expression analysis...... 43 2.3.4.1 Differential gene expression...... 45 2.3.4.2 Go terms and KEGG pathway...... 49 2.3.5 Differential splicing of isoforms...... 51 2.3.5.1 GO analysis...... 57 2.3.6 DNA methylation - Expression correlation...... 58 2.4 Discussion...... 61 2.5 Collaborative work statement...... 63

3 The neonicotinoid, Imidacloprid affects gut bacterial community in Bombus terrestris workers. 64 3.1 Introduction...... 65 3.2 Methods...... 71 3.2.1 Neonicotinoid toxicity test...... 71 3.2.2 Bee husbandry, imdicloprid exposure and gut sampling...... 71 3.2.3 DNA extraction and library preparation...... 72 3.2.4 Bioinformatic analysis...... 73 3.2.5 Taxonomy analysis...... 75 3.2.6 Differential abundance analysis...... 75 3.3 Results...... 77 3.3.1 Neonicotinoid anti bacterial toxicity...... 77 3.3.2 Bioinformatic metagenomic analysis...... 77 3.3.3 Diversity Indices...... 78 3.3.4 Differential abundance analysis...... 84 3.3.5 Taxonomy and relative abundance...... 87 Contents iv

3.3.6 Phylogenetic tree...... 89 3.4 Discussion...... 92 3.5 Collaborative work statement...... 95

4 The effect of neonicotinoids on geotaxis and the circadian clock in Drosophila melanogaster 96 4.1 Introduction...... 97 4.2 Materials and Methods...... 100 4.2.1 Negative geotaxis assay...... 100 4.2.2 Circadian activity...... 102 4.3 Results...... 103 4.3.1 Negative geotaxis climbing analysis...... 103 4.3.2 Circadian activity...... 104 4.3.3 Locomotor Activity Rhythms...... 107 4.4 Discussion...... 111 4.5 Collaborative work statement...... 114

5 The effect of black carbon on bacterial community of B. terrestris workers 115 5.1 Introduction...... 116 5.2 Materials and methods...... 119 5.2.1 Bee Husbandry and experimental design...... 119 5.2.2 Black Carbon preparation...... 119 5.2.3 Black carbon exposure and Faeces collection...... 119 5.2.4 Preparation culture media...... 120 5.2.5 Bacterial cultivation...... 120 5.2.6 Classification bacterial colony...... 120 5.2.7 DNA extraction...... 121 5.2.8 Amplification of 16s rRNA gene...... 121 5.2.9 Agarose gel electrophoresis...... 122 5.2.10 DNA sequencing preparation...... 123 5.2.11 PCR product purification...... 124 5.2.12 Data sequencing processing...... 124 5.2.13 Real time qPCR Quantification...... 124 5.2.13.1 DNA extraction, 16s rRNA amplification and purification. 124 5.2.13.2 Creating a gDNA standard curve...... 126 5.2.14 Experimental qPCR Setup...... 130 5.2.15 Quantification of total and specific gut phyla...... 131 5.2.16 Black carbon exposure, faecal collection and DNA extraction.... 132 5.2.17 16s rRNA gene sequencing...... 133 5.2.18 Bioinformatics analysis...... 133 5.3 Results...... 135 5.3.1 Bacterial isolation and identification...... 135 Contents v

5.3.2 CFU analysis...... 136 5.3.3 Quantiﬁcation of bacterial gut community...... 137 5.3.4 Bioinformatic metagenomic analysis...... 140 5.3.5 Taxonomy and relative abundance...... 140 5.3.5.1 Diversity Indices...... 144 5.3.6 Differential abundance analysis...... 147 5.4 Discussion...... 151 5.5 Collaborative work statement...... 154

6 Discussing ﬁndings 155 6.1 A summary of the chapters result...... 155 6.2 General discussion and conclusion...... 158

7 Future direction 163

A Supplementary information 165 List of Figures

1.1 Indirect effect of human activity to ﬂower pollination. Flowchart representing interactions between environmental stresses to ﬂower pollination (bumble bee) through indirect impacts...... 3

2.1 Examples of histogram of percentage methylation per cytosine. Histogram of % methylation per CpG (A), CHH (B) and CHG (C) for a control sample (C1). CpG % methylation shows a mild bimodal distribution with most sites modestly methylated and few fully methylated. Contrary to CpG, most CHH and CHG are modestly methylated. Both forward and reverse reads are reported...... 28 2.2 Methylated Cs distribution. Average proportion of methylation reads ± SD per CpG (A), CHH (B) and CHG (C) positions over genomic features. Control samples in black and neonicotinoid treated samples in grey.... 29 2.3 Shows proportion of differentially methylated Cs over genomic features for CpG (orange bar charts), CHH (red bar charts), CHG (grey bar charts). a CHG and CHH with p value 0.01 were used in the chart, regardles of < the % of difference between neonicotinoid and control samples...... 30 2.4 Methylation Exons switching. Methylation fold change for control (red) and Neo (blue) samples per exons in five genes. Exon usage fold changes are calculated based on the coefficients of a GLM fit. For each gene the intron exon annotation for each transcripts is shown...... 33 2.5 A KEGG pathway diagram shows the relationships of genes differentially methylated and map pathway of mitogen-activated protein kinase (MAPK) signaling cascade pathway : bter04013. A green rectangles show functional genes common of gene product (an enzyme), Red outline shows pseudogene formation and white rectangle shows absence gene...... 36 2.6 GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing significant differentially < methylated genes at CpG sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment.... 37

vi List of Figures vii

2.7 GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing significant differentially < methylated genes at CHHs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment.. 38 2.8 GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing significant differentially < methylated genes at CHGs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment.... 39 2.9 GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes containing significant < differentially methylated genes at CpG sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment...... 40 2.10 GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes containing significant < differentially methylated genes at CHHs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment...... 41 2.11 GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes containing significant < differentially methylated genes at CHGs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment...... 42 2.12 PCA plots showing correlation between samples and distance between neonicotinoid and control condition. Red dots representing imidacloprid samples and blue points representing control samples...... 44 2.13 Scatter plot showing log2 counts for each gene between sample 7 and 6 in control samples. The black line represents the real relationship between the x and values. The red line shows how extreme a value would need to be considered as an artefact...... 45 2.14 Heatmap tree showing 60 genes. 30 genes that are up regulated and 30 genes are down regulated. The yellow means high level of gene expression while blue shows reduced relative gene expression. Neo; neonicotinoid and ctrl; control...... 47 2.15 Volcano plot showing result data with log fold change, The x axis is the fold change (in log2 scale); the y axis is p-value (in log10 scale). Each dot on the plot is a single gene. Colour coding is based on adjusted p value, black dots genes p value greater than 0.1 and red dots genes with p value less than 0.1...... 48 List of Figures viii

2.16 GO term enrichment for Biological Processes. Enriched BP for GO terms (p <0.05) associated with genes containing significant differentially expressed genes. These rectangles are joined into different coloured ‘superclusters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment...... 50 2.17 Distribution of FPKM values across imidacloprid and control. samples from the same condition are shown in the same color: imidacloprid samples in blue, and control samples in orange...... 51 2.18 These plots show levels transcript structures of gene profile between neonicotinod and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow. A; LOC100642734 gene (facilitated trehalose transporter Tret1-2 homolog) and B; LOC100651516 gene neuferricin or Cytochrome b5 domain-containing protein 2 homolog 53 2.19 These plots show levels of transcript structures of the differentially expressed isoforms LOC100646534 gene profile between imidacloprid and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow...... 54 2.20 These plots show levels transcript structures of the of a gene profile between neonicotinod and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow. A; LOC100646534 L-asparaginase (L-asparaginase) and B; LOC100644045 gene (methyltransferase-like protein)...... 56 2.21 GO term enrichment for molecular function for GO terms (p 0.05) af- < ter FDR correction, associated to genes containing significant differentially alternative splicing events. Each rectangle represents a single cluster of closely related GO terms. These rectangles are joined into different coloured ‘superclusters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment.... 57 2.22 Average percentage of methylated CpG (A), CHG (B) or CHH (C) per gene. Control samples are in grey and Neo treated samples in white. Differen- tially expressed genes (DEG) and non differentially expressed genes (non- DEG) are plotted separately. Dots represent genes...... 59 2.23 The proportion of methylated CpGs (A), CHHs (B) and CHGs (C) is plotted against gene expression rank. One hundred “bins” of progressively increasing level of expression were generated and genes with similar level of expression have been grouped in the same bin. Solid lines represent control samples and dotted lines imidacloprid treated samples. The grey shading represents 95% confidence intervals...... 60

3.1 The digestive tract structure of honey bee and bumble bee species with bacterial localization in the different compartments: the crop, midgut, pylorus, ileum and rectum ( ﬁgure taken from Kwong and Moran [1])... 66 List of Figures ix

3.2 Illustration of nine hypervariable region in 16s rRNA gene were identified among bacteria. Red sections are start sites of primers (consiver ; green and orange raws are name of primers, light green (blue) sections are length of PCR amplification (DNA sequence). The figure refers to 16S rRNA and 16S rRNA Gene in the EzBioCloud Help center (http://help. ezbiocloud.net/16s-rrna-and-16s-rrna-gene/)...... 69 3.3 Boxplots showing Faith’s phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of imidacloprid and control samples. X axis indicates samples; Y axis indicates proportional of similarity bacterial species between samples...... 79 3.4 Boxplots show measure of community evenness (alpha-diversity indexes) in total difference proportions of bacterial species present in bumble bees. X axis indicates samples; Y axis indicates proportional of similarity bacterial species between samples . In a site with low evenness indicates that a few species dominate in the sample...... 80 3.5 Boxplots showing Shannon phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of imidacloprid and control samples. X axis indicates control and imidacloprid groups; Y axis indicates Shannon indexes of similarity bacterial species between groups. 80 3.6 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points control samples; red points neonicotinod samples. Distances between samples were calculated using weighted UniFrac distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees...... 81 3.7 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points control samples; red points neonicotinod samples). Distances between samples were calculated using unweighted UniFrac distance matrix. Samples stay close to each other means that those samples have communities with very similar overall phylogenetic trees...... 82 3.8 3D PCoA Plot showing a sample-by-sample distance based on Jaccard distances in bacterial communities, showing separation of samples by sample type between control and neonicotinoid. Each point represents one of the samples (green points control samples; red points imidacloprid samples. Distances between samples were calculated based on presence/absence. Samples close to each other means that those samples have percentage species with very similar overall phylogenetic trees.... 83 List of Figures x

3.9 Heatmap of the log of the coefﬁcient p-values for each of the balances. The rows of the heatmap represent samples and the columns of the heatmap represent balances. The colour scale represents microbial abundances at p value linear regression on the balance by computing the log ratios between species. Each of the tips corresponds to a taxon, Y represent split the data into partitions, for example Y0 is partitions between long branches (sub trees) and short branches (sub trees). light red in Yo balances is numerators for each balance and dark red is denominators.... 85 3.10 Shows number of unique taxa (phylotype) in both numerator and denominator partitions...... 86 3.11 Boxplots showing balance taxa summary, x axis represents average the log ratio for each sample, at each balance, calculated the isometric log ratio transform ; y axis control and treated neonicotionoid samples..... 86 3.12 Taxonomic level plot based on the 16S amplicon sequencing, the top phylum and sub phylum, calculated according to total relative abundance across the sample set. x axis is sample - y axis is relative abundance. The minor phylum which them abundance less 1 % are no showing in plot. grey bars phylum Actinobacteria, green phylum Firmicutes, orange sub phylum Gammaprotobacteria, blue sub phylum Betaprotobacteria..... 88 3.13 Neighbor joining phylogenetic tree showing the diversity of bacteria in control groups (A) imidacloprid groups (B). Alphaprotoacteria = blue, Acti- nobacteria = brown, Gamma and beta protobacteria = red; Firmicutes = green...... 91

4.1 The plot show the effect of imidacloprid on climbing ability (negative geotaxis) of M1217 (blue plots) and Canton-S (red plots) strains. X axis, concentration of imidacloprid and Y axis, the climbing coefficient the final distribution in the six trails of the counter-current apparatus. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles)...... 104 4.2 The plot shows activity in 12h light condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles). 105 4.3 The plot shows activity in 12h dark condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles). 106 4.4 The plot shows activity in 5 days constant darkness condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles)...... 107 List of Figures xi

4.5 The plot shows free running period in constant darkness condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles)...... 108 4.6 A double-plotted actogram showing average locomotor activity for 32 files in both strains of Drosophila melanogaster. X axis indicate time of day 24 h (08:30 am to 08:30 pm); Y axis indicates amount activity in every 30 minute. Each row of actogram showing showing 48 activity, each day activity showing twice belong each other except first day. A first four rows showing 4 days activity in 24 h (12 light and 12 dark) and rest rows are average activity in full constant darkness...... 110

5.1 An example of absolute quantification using a Standard curve. liner graph generated from the Ct values of stock (300 000 copies), dilution 1 (30,000 copies), dilution2 (3,000 copies), and dilution 3 (300 copies). The graph shows the mean of the cycle threshold (Ct) value along the Y axis against log DNA concentration (copy number) along the x axis. The efficiency of the PCR (E) was taken into account by the equation E = 1.0e-1 /slope -1.. 129 5.2 Shows median count of colony forming unit in faeces of bees exposed to black carbon and control bees. X axis represents treatment condition and Y axis represent number of bacterial colonies...... 137 5.3 The ratio of total bacterial population in bees faeces between treatment and controls. A ratio below 1 indicates less bacterial population in control bees compared to those treated with black carbon...... 139 5.4 The ratio of bacterial population in faeces between treatments (black carbon and control) of non reproductive workers Bombus terrestris. Charts ratio less than 1 indicates there were less bacteria in treated compare with untreated bees and equal to 1 mean no difference...... 139 5.5 Taxonomic level plot based on the 16S amplicon sequencing, calculated according to total relative abundance across the sample set. x axis is sample - y axis is relative abundance. Con_Pre control bees before treated (sugar water), Cont_Post control bees after 2 days treated with sugar water, BC-Pre bees before expsure to black carbon and BC_Post bees after treated with black carbon. The minor phylum which them abundance less 1 % are no showing in plot. grey bars phylum Firmicutes, , brown phylum Actinobacteria, blue phylum Gammaprotobacteria, and yellow assigned as unidentified bacteria...... 143 5.6 Boxplots showing Faith’s phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of black carbon (pre and post) and control (pre and post) samples. X axis indicates groups; Y axis indicates proportional of similarity bacterial species between samples.144 List of Figures xii

5.7 Boxplots show measure of community evenness (alpha-diversity indexes) in total difference proportions of bacterial species present in bumble bees. X axis indicates groups; Y axis indicates proportional of similarity bacterial species between samples...... 145 5.8 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points black carbon pre; red points black carbon post; green points control pre; yellow points control post). Dis- tances between samples were calculated using weighted UniFrac distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees...... 146 5.9 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points black carbon pre; red points black carbon post; green points control pre; yellow points control post). Dis- tances between samples were calculated using Jaccard distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees...... 147 5.10 Heatmap display log of the coefﬁcient p-values for each of the balances. The rows of the heatmap represent samples and the columns of the heatmap represent balances. The colour scale represents microbial abundances at log ration of p value linear regression on the balance. Each of the tips corresponds to a species, Y represents splitting the data into partitions, for example Y0 is partitions between long branches (sub trees) and short branches (sub trees). light red in Yo balances is numerators for each balance and dark red is denominators...... 149 5.11 Shows number of unique taxa (phylotype) in both numerator and denominator partitions...... 150 5.12 Boxplots showing balance taxa summary, x axis represents average the log ratio for each group, at each balance, calculated the isometric log ratio transform ; y axis BC_Post: bees after treated black carbon, BC_Pre: bees before treated with black carbon, Con_Post; Bees before treated with sugar water, Con_Pre; bees after treated with sugar water...... 150

A.1 The captured image shows the code function of pipeline to remove outlier (artefact) genes between samples in same condition. The code was generated and provided by Sascha Otts group (they ran the RNASeq on the MIBTP)...... 166 List of Tables

2.1 Shows number of sequences analysed, number of sequences with a unique best alignment, statistics summarising the bisulﬁte strand the unique best alignments, number Cs and percentage methylation of cytosines in CpG, CHG or CHH context (where H can be either A, T or C)...... 27 2.2 Genes with differentially methylated Cs (p value less than 0.01) in multiple contests (CpG, CHH, CHG). * less than 5% differentially methylated CHG, p value less than 0.01...... 31

3.1 Shows summary of sequences were retained after trimming and removing low quality reads...... 78

4.1 illustrating analysis of locomotor activity data in full darkness condition from day 1 to 5 (DD1-5) and day 6 to 10 (DD6-10). N.ahr; number files did not perform stable rhythm; N.dead; number of files dead; N.CR; number of flies performed more than one thyme during 24 h; period , periodicity of rhythm locomotor activity and STD, standard division of amount of variation in periodicity between flies in same condition...... 109

5.1 Hot and cool cycler steps of PCR ampliﬁcation 16s rRNA gene using Kapa Taq Polymerase...... 122 5.2 showing stages of PCR process...... 123 5.3 shows calculation the mass of gDNA (16s rRNA ) containing the copy #s of interest...... 127 5.4 shows the concentrations of gDNA needed to achieve the copy#s of interest...... 127 5.5 showing the calculated volumes of gDNA and diluent for all 5 dilutions.. 128 5.6 showing the sequence and amplicon size of the primers used in qPCR approach to look for broad changes in bacterial species after exposure to black carbon...... 130 5.7 Shows bacterial species which were isolated from faeces B. terrestris colonies On BHI media. Blasting query cover For species in NCBI database were 100 % ...... 136

xiii List of Tables xiv

A.1 These tables show loc number (gene ID), transcripts and names for genes that deferentially methylated Cs (p value less than 0.01) in multiple contests (CpG, CHH, CHG) between treated and untreated bees. Diff Meth; less than 5% deferentially methylated in CHG, CHH sites and 10% in CpG sites...... 167 A.2 These tables show gene ontology terms and biological process for genes that differentially methylated Cs in multiple contests (CpG, CHH, CHG) between treated and treated bees...... 183 A.3 These tables show genes differentially expressed between treated and untreated bees with name of transcript for each gene...... 218 A.4 These tables show gene ontology terms and biological processes for genes that differentially expressed between treated and untreated bees...... 248 A.5 These tables show gene ID and description for genes that deferentially expressed transcripts between treated and untreated bees...... 265 A.6 These tables show GO terms and biological processes for genes that differentially expressed transcripts between treated and untreated bees.... 267 Abbreviations

AAdenine ANOVA Analysis of Variance BGI Beijing Genomics Institute BS-seq Bisulphite sequencing bp Base pairs (of nucleic acid) CCytosine oC Degrees Centigrade CFU Colony Forming Unit CpG Cytosine phosphate Guanine sequence within DNA

ddH20 Double distilled water DEG Differentially expressed genes DEI Differentially expressed multi-isoform genes df Degrees of freedom DMR Differentially methylated region DNA Deoxyribonucleic acid DNMT DNA methyltransferase FDR False discovery rate FPKM Fragments Per Kilobase of transcript per Million mapped reads g grams GGuanine GC-content Guanine-Cytosine content GLM General linear model

xv Abbreviations xvi

GO Gene ontology logFC log fold-change miRNA micro RNA mm Millimetres µl Microlitres µm Micrometres

µM Micromoles mRNA messenger RNA NCBI National Center for Biotechnology Information ng Nanograms NOEC No Observed Effect Concentration OTU Operational Taxonomic Unit PBS Phosphate-buffered Saline PCA Principal component analysis PCoA Principal coordinates analysis PCR Polymerase chain reaction PERMANOVA Permutational Multivariate Analysis of Variance PNACL Protein Nucleic Acid Chemistry Laboratory ppb part per billion QIIME Quantitative Insights Into Microbial Ecology qPCR quantitative Polymerase Chain Reaction REVIGO Reduce and Visualise Gene Ontology RNA Ribonucleic acid RNAi RNA interference RNA-seq RNA sequencing SNP Single Nucleotide Polymorphism TThymine 16s rRNA 16S ribosomal RNA Chapter 1

General introduction

1.1 Interactions between plants-pollinators-pests-

ecotoxicology

Since Darwin (1859), biologists have been increasingly aware of the importance and complexity of mutualistic symbiosis between plants and animals which have multiple influences on community population, dynamics and diversity [1–4]. Insects pollinators are highly efficient at transporting pollen from one flower to another which helps in plant biodiversity and increases agriculture production [5]. In contrast, many of the known insect species are considered as agricultural pests which inflict damage to plants bee-friendly flowers. In 1947, Williams identified insect pests as any insect in the wrong place, based on the assumption that insect attacks cause damage to the plant [6,7]. Different procedures and methods are used to control harmful insects, one of the most popular is chemical methods [8]. The effects of these chemicals have been studied by many researchers in the areas of toxicology, ecotoxicology (atmospheric, aquatic and soil chemistry) and agriculture [5]. These agrochemicals have direct and indirect effects on non target organisms, and their impacts on the structure and functionality of natural ecosystems affect long-term sustainability [9]. All pesticides are toxic to a

1 Chapter 1. General introduction 2 greater or lesser degree to all animals, so their release in the environment could not be without risks to other organisms.

Insecticides are not the only man-made contaminants released into the pollinators’ environment, because bees are mobile and sources of food are not protected from industrial chemicals, mining wastes, pharmaceutical residues and the large list of pollutants that humanity produces. For example the burning of fossil fuels or biomass produces huge amount of black carbon to the environment [10]. Surveys such as that conducted between 2009 - 2016 by Singh et al. have shown that the concentration of black carbon is 15 PM2.5/µg m-3 across the UK [11]. The Air Quality Expert Group reported that there are significant amount of black carbon in both suburban and village sites in the UK [12]. However, urban habitats can contain remarkably high pollinator species, particularly bumble bee. Many studies have found a decrease in the species richness of pollinating insects with increased urbanization [13]. Bees foraging and pollination efficiency decreases with increasing air pollution [14]. Fine particulate matter were detected in bees bodies after foraging performance [15].The interaction between bees and the environment are summarized in the flowchart (1.1). Chapter 1. General introduction 3

FIGURE 1.1: Indirect effect of human activity to ﬂower pollination. Flowchart representing interactions between environmental stresses to ﬂower pollination (bumble bee) through indirect impacts.

1.2 Bumble bees and their importance in nature

What are bumble bees? Extant bumble bees are members of the genus Bombus and along with the related honey bees are belong to the family Apidae, Order Hymenoptera, class Hexapoda (Insecta). There are 250 species of bumble bee in the world, with 27 species being recorded in the United Kingdom, and only 7-8 sub species considered widespread and abundant [16]. The most common species widely distributed across England, Wales, Northern Ireland and southern and central Scotland is the buff-tailed Chapter 1. General introduction 4 bumble bee, Bombus terrestris audax [17]. The majority of bumble bees are social insects, they are often described as ‘primitively eusocial’,living together in a colony with a single queen, dozens of non-reproductive workers and few males at the end of colony life. The main function of the male is to mate with virgin queens. Workers are responsible for doing all the work inside and outside the nest, including collecting and processing food, caring for the raising new offspring, defending the colony, building honey pots, and maintaining the colony’s temperature. The colony size is small and individuals fairly large compared with the majority of bees species [18, 19].

Economic value Bumble bees are important pollinators for many agriculture crops including tomatoes, peppers, and blueberries [16, 20,21]. About 75% of crop species grown for human consumption and livestock rely on animal mediated pollination [22]. In general, natural pollination services of wild bees has been valued at over £430 million a year in the UK [23]. Bumble bee services have been estimated at $3 billion a year in the United States crop production, being over $750 million in soybean production alone [24]. Bumblebees are much more efﬁcient pollinators than honeybees. Because, they are robust, large in girth, have more hairs on their body and many species have longer tongues [17]. Also honey bees are unable to perform the “buzz pollination” required by tomatoes while bumble bees able to vibrate tomato plant at around 400 Hz [25]. Because bumble bee play important role in crop production, therefore bumble bees are required in worldwide and importing into non native country. Global trade in bumble bee colonies has been estimated at over one million nests per year. An estimated 50,000 colonies of bumblebees are imported to United Kingdom each year [17], and Japan alone imports about 40,000 bumble bee colonies annually [26]. Bee pollinators also facilitate the maintenance of sexual reproduction of crops and wild ﬂowers by carrying pollen from one plant to another [27].

Wild pollination and biodiversity Bumble bees are one of the largest groups of wild ﬂower-visiting insects and they are fully depend on pollen and nectar throughout their Chapter 1. General introduction 5 lives [28]. They play a key role in the maintenance of plant biodiversity and the survival of plant species, through movements of genes between populations [29]. In fact, the survival numerous wild ﬂower plants depends on bumble bee for pollination and take away the bumble bee from nature, both wild plants and animals which depend on them will be threatened. So without bumble bee wild life become less colourful and less interesting place.

1.3 Evidence for the decline in bee populations

Over the last few years, the decline in populations of both bumble bees and honey bees have reached a critical point. A large and growing body of studies have investigated the high rate of honey bee colony collapse, in a worldwide phenomenon known as colony collapse disorder (CCD) [30, 31]. Data from a survey in the USA showed that managed colony numbers have declined by 45% from 2012 to 2013 [32], In Belgium, Italy, Netherlands, United kingdom and Sweden and Ireland, the loss of honey bees colony exceeded 25% [33,34]. There is much evidence that bumble bee populations are also under threat and declining globally. United States Fish and Wild Life service reported, that before 1990 in the United States, the bumble bee was common and some species were in high abundance across 28 states from Connecticut to South Dakota [35]. While today, abundance has decreased dramatically in 13 states and the loss was estimated above 87 percent with colonies living in fragmented, small populations [35]. In North America, both Bombus terricola and Bombus occidenttalis have severely declined since the mid 1990s and have undergone substantial range contraction and local extinction [36, 37]. Bombus frankini was one of the most abundance species in California and Oregon but since 2006 has not been recorded [38]. In chain, decline and extinct some species of bumble bee have been recorded since the early 1990s [39]. Four species of bumble bee (B. armeniacus, B. cullumanus, B. serrisquama, and B. sidemii) were for- merly abundant and widespread throughout the entire region of Europe. Now they are endangered and may go extinct in the near future [40]. Three species of bumble bees Chapter 1. General introduction 6 have gone extinct across the UK since World War II [38]. A hypothesises for the rapid decline and crises in bee that has been drawing increasing attention, are habitat loss, climate change and pollution [41, 42].

1.4 Factors affecting insects communities

A wide range of anthropogenic environmental changes have been mentioned in the media and academic research as possible causal agents of bees decline [41, 43, 44]. These factors include; pests and diseases [41, 45], diet and nutrition [46], habitat loss [47], air pollution [48] and exposure to pesticides [49]. Most of these may be interre- lated to each other and in combination increase stress on bee populations [50–52]. One factor, widely documented is insecticide use - in particular neonicotinoids [51,53–55]. It has conclusively been shown that they have negatively inﬂuenced bee populations. In same line, the data from different researches revealed a dramatic decline and extinct on some species of bees since 1990s and this period starting neonicotinoids production and using them on the farm. So, I hypothesize that some of these large array of effects could be mediated by neonicotinoids and air pollution which affecting bee health. Here, I began my study of the effect of neonicotinoid insecticides and Black carbon (air pollution) on epigenetic and gut microbiota in the bumble bee.

1.4.1 Black carbon

Black carbon is a ingredient of fine particulate matter, the sources in the atmosphere come through incomplete combustion of fossil fuels, biofuel, biomass and exhaust fumes of road traffic [56]. The particle sizes are less than 2.5 µm, easily transported through the air and can suspended in the atmosphere for about one week [57]. Black carbon introduces to the ocean via the rivers [58] and rain to the soil [59]. It has been found in the terrestrial ecosystem and identified in the form of polycyclic aromatics [60]. Chapter 1. General introduction 7

There is evidence that black carbon has multiple impacts on the environment, inﬂu- encing changes in patterns of rain and clouds, and alters the melting of snow and ice cover [61]. Black carbon can cause health problems in humans [62], including cardiac and respiratory morbidity [63]. Residues in the soil increase cation exchange capacity, biogeochemical processes and nutrient retention and recycling [64]. Black carbon negatively affects microbial communities residing in the phyllosphere and rhizosphere in the soil [65]. Currently, a group in lab 121, Departement of Genetic and Genome Biol- ogy are looking at the effect of black carbon on microbiota in mice model and bumble bees. Recently Hussey et al. showed that black carbon can drastically changes the development of bacterial bioﬁlms [66]. It could be seen these broad mode of action may have a serious effect on microbial communities in the organism which expose to black carbon. Such as foraging bumble bees are potentially heavily exposed to black carbon, particularly in urban areas.

1.4.2 What are neonicotinoids?

Neonicotinoids are a family of neurotoxc pesticides whose chemical structure is similar to nicotine. They bind to and activate the nicotinic acetylcholine receptor in the insect central nervous system (particularly the α4β2 subtype) [67, 68]. α4β2 is implicated in learning and memory [69]. The neonicotinoids act predominantly on Kenyon cells that are the major neuronal cell type in the mushroom bodies of the bee brain [70]. Neon- icotinoids disrupt the function of insect neurons and cause paralysis and death [71]. The toxic effect of the neonicotinoids are based on their strong against neurotransmitter, overstimulating neurotransmission and effectively halting nerve function, and they bind to the nicotinic acetylcholine receptor at the location of acetylcholine. Neon- icotinoids act selectively to the insects nicotinic acetylcholine receptor while are made to have no or low acute and chronic toxicity to mammalian species, as well as to ﬁsh or birds [68]. This due to differential sensitivity of the insect and vertebrate nicotinic Chapter 1. General introduction 8 acetylcholine receptor subtypes [72]. In comparison with other pesticides, all neonicotinoids types are a much lower risk to non-target organisms and to the environment.

Neonicotinoids have become one of the most important types of pesticides in the world which play a crucial role in modern crop protection [73]. They may also be used in controlling fleas and lice on dogs and cats and are applied in non agricultural applications, such as on the lawn or garden, or to control pests including termites, turf pests, and ants [74,75]. Neonicotinoids are particularly used for controlling those insects that have chewing and sucking type mouth parts such as aphids, whiteflies, leaf-and plant hoppers, thrips, some micro-Lepidoptera, and a number of Coleopteran pests [76]. Neonicotinoids work inside plant systemically, meaning they are absorbed by the plant and transported to all tissues where they remain active for many weeks [77]. However, neonicotinods’ tissue distribution and stability, which is important for their efficacy, determines their lack of species-specificity and their negative effects on the environment [78]. The neonicotinoids family includes acetamiprid, clothainidin, nitenpryam, nitthiazine, thiacloprid, thiamethoxam and imidacloprid. They have been registered in 120 countries as effective tools to control a variety of pests and cover approximately 25% of the global insecticide market [74]. By 2009, in United states their production and marketing grew rapidly valued at $2.63 billion and neonicotinoid Imidacloprid accounted for the greatest proportion (41.5 %) of this marketing value [74]. In 2011, the UK, 91% of seed dressing accounted were neonicotinoid imidaclopird [79].

The negative effect of neonicotinoids on honeybees, wild bees and other pollinators have been the subject of intense debate within the scientiﬁc community. More recently a report from the European Food Safety Authority (EFSA) about risks of neonicotinoids to bees has conﬁrmed the risks [80]. The Authority has updated its risk assessments of three neonicotinoids – clothianidin, imidacloprid and thiamethoxam and In 2013, the European Union imposed a partial restriction on these three types of neonicotinoid [81]. The National Farmers Union regularly requests and is granted licenses for emergency use of neonicotinoids [81]. Outside of Europe, few countries have introduced Chapter 1. General introduction 9 restrictions on neonicotinoid use and they remain the most widely used insecticides in the world.

1.4.2.1 Imidacloprid and environmental residues

Imidacloprid is produced in many different forms including liquids, granules and dusts and comes as colourless crystals with a weak characteristic odor, high water solubility, which is an important characteristic for agricultural purposes, the molecular formula C9H10ClN5O2, a molecular weight 255.7, molar mass 255.66 g/mole [82]. Imidaclopird is applied to control insects on grains, maize, fruits, vegetables and potatoes [83]. Ac- cording to report by Food and Environment Research Agency (fera), From 1990 to 2015 imidacloprid was most widely used in the United Kingdom. In 2013, 36,795 hectors (ha) were treated, in 2014 the area were 21,410 ha while in 2015 estimated by 20,981 ha and in 2016 the area decreased to 2,069 ha [84]. One of the main features about imidacloprid is that has low vapor pressure under natural ﬁeld condition and systematically transfers from roots to all parts of the plant which exist in all parts of a plant. So that, it could be exists in leaves, ﬂower, pollen and nectar [85].

Imidaclopird is moderately persistent compound in the environment and remains in the soil for several months [86]. Imidacloprid has been detected in aquatic sediments and water [87]. To give an example: British rivers are heavily contaminated with neonicotinoids [88]. In nectar and pollen, the range of neonicotinoids concentration is in parts per billion (ppb) [89]. In an experiment in greenhouse, the sunﬂower seeds were treated by imidacloprid and the seeds were dressed with imidacloprid at 0.7 mg per seed. Analysis of seeds showed that the residues of imidacloprid are in a range about 3.9 ppb pollen and 1.9 ppb in nectar but were found in 85% of the samples [90]. An- other study in 2003 showed an increasing the imidaclopird residues in sunﬂowers with a ranges between 5 to 10 pbb [91]. Imidacloprid concentration is different between leaves from up to down [92]. Recently, researchers found that 25 % of British honey Chapter 1. General introduction 10 was contaminated with bee-harming neonicotinoids [93], 75% of global honey samples across the world contaminated with cocktail of neonicotinoid classes [94].

1.4.2.2 Lethal effect of neonicotinoids

The acute toxicity of neonicotinoids are low to bees compared with older families of insecticides such as organophosphorus, pyrethroids, carbamates [95,96]. Laboratory studies calculated that LD50 (dose that kills 50% of individuals) values of imidacloprid is 18 ng/bee in the honey bee, Apis mellifera [97]. Another study estimated the acute toxicity value of imidacloprid at 48 h in about 5 ng/bee for both A. m. mellifera and A. m. caucasica [96]. While another study showed LD50 value of imidacloprid in about 60 ng/bee at 48 h [98]. In 1994, Stark‘s group exposed three species of bees ( Apis mellifera, Megachile rotundata and Nomia melanderi ), the bees were equally susceptible to imidacloprid (24-h LD50 0.04 ng/bee) [95]. Abnormal behaviours (restless, apathetic, trembling and falling over) with mortality rates up to 45% were detected when honey bees were exposed to syrup contaminated with 125 µg/kg of imidacloprid insecticides [99]. On the other hand, a concentration of 10 µg/kg is considered to be “No Observed Effect Concentration (NOEC)” on bees [100]. But, in 2014, Thompson et al. showed that, the NOEC in the diet varies between 20 and 40 ug/l in the bumble bees Bombus terrestris, this was calculated in a ten-day experiments. The range of ﬁeld-realistic concentrations (5 to 10 ppb) of neonicotinods clothianidin is the non lethal dose against larvae of solitary bee Osmia bicornis [102]. 100 µg/L of clothianidin causes 100% mortality of bumblebee Bombus terrestris while no reduction in food intake was recorded at concentration 1 µg/L but reduction in consumption was noted at 10 µg/l [101]. It appears that there are high discrepancy in acute toxicity in bee group and this underlines the assumption that other factors age, species, caste, colony, neonicotinoids classes) can change value of LD50, or may due to differences in the experimental methodolo- gies. Chapter 1. General introduction 11

1.4.2.3 Sub-lethal effects of neonicotinoids

There is evidence that sub-lethal doses of neonicotinoids affect physiological states, memory and behavioural in beneﬁcial insects like honey bees and bumble bees. This severely reduces their ecological performance and negatively affecting their population dynamics [53, 100, 103].

Chronic Toxicity is long term exposure of an organism to a toxicant or other stressor which causes adverse effects in on bees health including changes in growth, reproduction and longevity [104]. Nurse and larvae of bees may be exposed to the neonicotinoids as foragers collect potentially contaminated pollen and nectar to be stored inside the hive for the long time [105]. For example, an overall, signiﬁcant increase of mortality of queens of Bombus impatiens Cresson was detected during the 6 weeks treatments contain 20 ppb of imidacloprid in the greenhouse, this with including adult mortality, brood development, colony weight, food consumption and foraging efﬁciency of adults [106]. Sub-lethal effects of dietary neonicotinoids insecticide exposure can reduce queen fecundity and colony development of honeybee [105]. Moreover, fed with syrup and pollen containing imidaloprid at 10 and 6 ppb causes a reduction of the number of larvae produced in colonies of Bombus terrestris [107]. Combinations of neonicotinoids and pyrethroid at sub-lethal dose increases bumble bee workers mortality, reduces brood development and increases the propensity of colonies to fail [108].

Behavioural and learning effects in the many studies, the behavioral effects of neonicotinoid insecticides have been shown over the last 20 years. However, there is considerable discrepancies among bee species in foraging behaviour and division of labour in nurse workers. which may cause the effects of pesticides vary from one species to others and between colony individuals [109]. Effect of neonicotinoids on foraging, homing, mobility, navigation and locomotion, and pollen collecting have been highly documented in the honey bee [99, 110–112]. For example, imidacloprid at 20 ppb decreases Chapter 1. General introduction 12 the foraging activity in honey bee [113, 114]. A recent meta-analysis based on laboratory and field studies highlighted that, the honey bee is generally very sensitive to neonicotinoids compare to the bumblebee [109]. Despite this, there is some evidence that non lethal doses of neonicotinoids may reduce foraging efficiency of bumble bee as well [115]. 5.32 ppb of neonicotinoids for two weeks cause a reduction in feeding in two out of these four species Bombus terrestris, B. lucorum, B. pratorum and B. pas- cuorum [116]. In the course of collecting nectar and pollen by the bumble bee and honey bee, floral parameters such as location, shape, color and odor of flowers are associated to a bees learning process [117]. The memory of bumble bees is significantly impaired following chronic exposure to neonicotinoids [112]. Control groups showed a higher learning level than those exposed to both 10 ppb and 25 ppb of neonicotinoids treatment [118]. Non-lethal doses of imidacloprid impair facilitation of cognitive performance in the honey bee [119, 120]. Some other studies have also shown that forg- ers of bumblebee become less efficiently for pollen collection when exposed to imidacloprid [108, 121, 122]. It would be useful to investigate whether slower learning may negatively affect the dynamics and efficiency of non neural parts such as reproductive system, digestive system and gut microbiota.

Non neural effects Alarmingly, when it comes to side-effect of neonicotinoid pesticides on non neural parts, the news is no better. Field realistic concentrations of imidacloprid can severely affect the hemocyte density, encapsulation response, and antimi- crobial activity of western honey bees Apis mellifera L.[123]. Neonicotinoid pesticides and nutritional stress synergistically increase terhalse sugar levels and then lead to reduced survival in honey bees [124]. There is a ﬁnding that ﬁeld realistic dose, sub-lethal doses of imidacloprid decreases sperm viability, dysfunction reproductive anatomy (ovaries) and physiology state ( spermathecal-stored sperm quality and quantity) in the queens of honey bee [125]. Another study showed that neonicotinoids alter morphology and physiology in the ovary of honey bee [126]. Subsequent research found that Chapter 1. General introduction 13 two weeks of exposure to the higher concentration of neonicotinoids caused a reduction in the average length of terminal oocytes ovary development in multiple species of wild bumble bee queens [116]. Imidacloprid decreases the size of hypopharyngeal glands in the honey bee [127]. Moreover, in 2017, Ciereszko et al. demonstrated that chronic imidacloprid dietary exposure has severe detrimental effects on motility, viability and mitochondrial membrane potential of sperms in drones of honey bee. In bird, residues of imidacloprid were detected in the liver of affected pigeons [129]. Im- idacloprid exposure cause decrease Glutathione and elevated levels of thiobarbituric acid reactive substances (TBARS) in liver tissue [130]. Accumulating evidence suggests that neonicotinoids may have long-term adverse effects on the digestive system. For example, exposing Africanized honey bees, Apis mellifera, to a sub-lethal dose of thiamethoxam lead to ultrastructural changes in the midgut and Malpighian tubules and this damage greatly affects the absorption and excretory functions [131].

Gut bacteria effect Up to now, far too little attention has been paid to look at the effect of neonicotinoids on gut microbiota. But, there are several lines of evidence suggesting an effect of neonicotinoids on gut microbime either directly or indirectly. The neurobiological basis of effect neonicotinoid on bees is strongly approved [99, 110– 112]. Generally, there is a strong link between gut microbiome and function of the central nervous system [132]. Other pesticides, for example, the fungicide Pristine adversely affects the relative abundance of Lactobacillus sp. Firm 4 and Firm 5 in the guts of honey bee colonies [133]. The composition of the gut microbiome in bees is very important in many biochemical process. The gut microbiom is very sensitive and it can be inﬂuenced or disturbed during the course of physiological and behavioural changes [1, 133]. Exposure to neonicotinoid has been shown to be related to a variety of ongoing complications in the bees body, as well as negatively affected locomotion, behavior, learning and memory, olfactory performance, and foraging behavior of bees [134–136]. One longitudinal study with different environmental landscape of oilseed rape found that potential effect of exposure neonicotinoid insecticides, on key honey Chapter 1. General introduction 14 bee gut bacteria [137]. Studying the relationship between gut bacteria and neoincotinoids will help us understand the deeper connections between neonicotinoids exposure and the health of bees.

1.4.3 Epigenetic effects

There is much evidence that the external environment’s effects upon genes can influence phenotype, and some of these effects can be inherited in to many generation [138]. Epigenetics can play a role in the interplay between man- made chemicals and natural ecosystems, and their constituent species [138]. Epigenetics is the study of heritable changes in gene expression that do not involve alterations in the DNA sequence. Examples of mechanisms include DNA methylation, histone modifications and microRNA expression [139]. These molecular modifications can be triggered by environmental factors [140]. Based upon this, I hypothesize that there is a strong possibility that neonicotinoids have an effect on the methylation status of the bees. Thus, understanding the epigenetic mechanisms fills a current knowledge gap in the field of molecular ecology effects of neonicotinoids on the bumble bee that eventually serve as indicter for neurotoxic pesticides to the wild and manged pollinators.

DNA Methylation, is an epigenetic marker involving in many biological function in bees developmental processes, the control of reproductive status, memory, behaviour and longevity [141]. All these important factors have shown to be affected by neonicotinoids [134–136]. In mice, nicotine decreases the expression of DNA-methyltransferase 1 and reduces DNA methylation in cells expressing nicotinic but not muscarinic AChRs, suggesting epigenetic consequences following repeated activation of nAChRs [142]. Exposure to low dose of dichlorodiphenyltrichloroethane (DDT) alters the methylation pattern in the hypothalamus of young male rats [143]. With Asian tiger mosquito Ae. albopictus global DNA methylation alters have been shown to be involved in the regulation of decreasing insecticide sensitivity towards the neonicotinoid imidacloprid on Chapter 1. General introduction 15 the phenotype level [144]. Organochlorine pesticides aslo showed the effect on DNA methylation in prepubertal female Sprague-Dawley rats [145].

The effects of pesticides on the DNA methylation can be attributed also to a change in the miRNA expression profile, thus leading to changes in gene regulation [146]. More- over, methylation level was positively associated with gene expression for genes that were differentially methylated in male and queen castes of Solenopsis invicta fire ants [147]. In fruit fly, an increase of oxidative and nitro-reduced metabolites were detected during metabolism of imidalcoprid [148]. Oxidise stress resistance of silverleaf whitefly Bemisia tabaci to imidacloprid pesticides was detected by over expression of CYP6CM1 gene during exposure [149]. The molecular genetics of insecticide resistance has confirmed the relative importance of cytochrome P450s enzymes in metabolic resistance [150].

1.5 Aim and justiﬁcation

The overall aim of this thesis is to quantify the effect of neonicotinoid imidacloprid on DNA methylation, gene expression, alternative splicing, behaviour and gut bacterial community and the effect of black carbon on the gut bacterial community of buff tailed bumblebee, Bombus terrestris audax.

The speciﬁc objectives were to:

1. In chapter2, I analyse three whole methylome (BS-seq) libraries of the brains of neonicotinoid imidacloprid exposed workers and three libraries from unexposed, control workers in order to elucidate neonicotinoids’ effects on the methylation status of bumble bee workers Bombus terrestris audax. I also analyse sixteen transcriptome (RNA-seq) libraries (9 control vs 7 imidacloprid) in order to discover the gene expression changes and alternative splicing associated with imdiacloprid exposure. Chapter 1. General introduction 16

2. In chapter3, I examine whether the composition of the gut bacterial community of non reproductive workers Bombus terrestris audax. is affected by imidacloprid. The comparative is based bacterial community profiling, using a deep sequencing of the V4 region of the 16S rRNA gene, fifteen libraries of imidacloprid exposed workers and fifteen libraries from control workers.

3. In chapter4, I used Drosophila melanogaster as a model for the effects of imidacloprid on bees. This were performed by looking on geotaxis and the circadian clock fruit ﬂy during exposure to food contain ﬁve different concentration of imidacloprid (0 ppb, 2.5 ppb, 5 ppb, 10 ppb and 20 ppb).

4. In chapter 5, I set out to investigate the hypothesis of the effect of black carbon on the stable gut bacterial community abundance and composition of non reproductive workers Bombus terrestris audax. We explore this possible differences between workers exposed to black carbon and unexposed, using culture based methods, absolute quantiﬁcation method by performing simple quantitative polymerase chain reaction (RT-qPCR) approach and deep ampliﬁcation sequencing of a region (V4) of 16S rRNA gene using Ion Torrent sequencing technology. Chapter 2

The neonicotinoid, imidicloprid affects gene expression, alternative splicing and DNA methylation in Bombus terrestris

17 Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 18

2.1 Introduction

Neonicotinoids are effective insecticides used on many important arable and horticul- tural crops, most frequently as seed dressing [73]. Imidacloprid is one type of neonicotinoid and it has been widely used throughout the world since the 1990s [44, 84]. Laboratory and field studies have shown that neonicotinoids disrupt the function of insect neurons causing paralysis and death [71]. However there are a large number of sun-lethel effects including negative effects on mortality, foraging efficiency, flying ability, mobility, navigation and locomotion, and pollen collecting [99, 101, 115, 124]. For more details see section 1.4.2 in chapter 1.

I hypothesize that some of these large array of effects could be mediated by neonicotinoids affecting the epigenetics of bees. Epigenetics is defined as a heritable change in gene expression without any change in the DNA sequence [139]. Epigenetics play a vital role in understanding the relationship between the genotype and phenotype and this relationship has recently been documented as a major challenge for molecular ecology [151]. Environmental contaminants have been found to affect the epigenetics of a diverse range of animal species from water fleas to polar bears [152] and include metals, endocrine disrupting compounds, air pollution, persistant organic pollutants and pesticides [153]. If environmental contaminants affecting the epigenetics of organisms is common, this would have major effects on the field of ecotoxicology. This field is based on the idea that exposure and effects can be directly linked. Epigenetic modes of action would break this direct link [152]. This would mean risk assessments would have to take into consideration that the toxic response may affect future generations. Epigenetics may also be able to explain well established examples of persistent and multigenerational responses to environmental contaminants [152].

Methylation, an epigenetic marker involving the addition of a methyl group to a cytosine. In bees, DNA methylation is mainly located at the cytosine in CpG dinucleotides, although a small proportion of methylcytosines can be detected in the context of CHG or CHH (H stands for A, G, or T) [154–158]. Methylation of DNA is catalyzed by DNA Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 19 methyltransferase 1 and DNA methyltransferase 3 [159]. Social insects have multiple copies of DNA methyltransferase 1 which plays an important role in maintenance DNA methyltransferase during DNA replication [154, 160]. Inhibition of DNA methyl- transferases in honey bees (Apis mellifera) affects whether a larva becomes a queen or worker [160]. 561 differentially methylated genes are involved in caste determination between adult queen and worker of honey bee [160]. Methylation differences between of the genome reproductive workers and non-reproductive workers [161]. Social epigenetic studies in social insects showed that DNA methylation has important effects on memory, longevity, social behaviour and behavioural plasticity [141, 162, 162], all these features have shown to be affected by neonicotinoids [118, 125, 163]. In mammals, methylation on gene promoters leads to a reduction in gene expression [164].

Most hymenopteran insects predominantly display methylation found in coding regions [154]. In bees methylation status are targeted to “gene bodies” or transcriptional units rather than nongenic regions [165–167]. Gene body methylation can be associated with active transcription and differential gene splicing [160, 168]. There are many evidence that alternatively spliced genes may play important roles in developmental plasticity in many species [169, 170]. This suggests a link between gene body methylation, gene expression and alternative splicing in the bumble bees as result of neonicotinoid.

In mice, nicotine decreases the expression of DNA-methyltransferase 1 and reduces DNA methylation in cells expressing nicotinic but not muscarinic AChRs, suggesting epigenetic consequences following repeated activation of nAChRs [142]. Global DNA methylation changes have been shown to be involved in the regulation of decreasing insecticide sensitivity towards the neonicotinoid imidacloprid on the phenotypic level [144].

As epigenetics is deﬁned as a change in gene expression without any underlying change in DNA sequence, I will also directly measure gene expression. Gene expression differences due to neonicotinoid exposure have been found in honey bee larval workers, Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 20 adult workers and queens [171–173].

However, changes in a gene’s overall expression are not the only way expression can be affected. For example, the difference between head lice (harmless) and body lice, (disease vector) is caused by the alternative splicing of an identical genome not by differential gene expression [174]. Alternative splicing results in a single gene coding for multiple proteins because exons are either included or excluded from the ﬁnal, processed messenger RNA (mRNA). I am unaware of any studies which have looked at the effects of neonicotinoids on alternative splicing in insects.

In this study I analysed whole methylome (BS-seq) libraries of the brains of twelve neonicotinoid exposed workers and twelve control workers from each of three colonies in order to elucidate neonicotinoids’ effects on the methylation status of bumblebee workers. I also analysed transcriptome (RNA-seq) libraries form other three colonies in order to discover the gene expression changes and alternative splicing associated with neonicotinoid imidacloprid exposure. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 21

2.2 Materials and methods

2.2.1 Bee husbandry

Six colonies of Bombus terrestris audax were purchased from Agralan, UK. Each colony contained a queen and on average ten workers and a small amount of brood. They were kept in wooden nest boxes and maintained under red light at 26◦C and 60% humidity on a diet of 50% v/v apiary solution (Meliose-Roquette, France) and pollen (Percie du set, France) [161]. Three colonies were used for the RNA-seq experiment and the other three for the BS-seq experiment.

2.2.2 Neonicotinoid feeding and brain sampling

Groups of 5 callow workers born on the same day were reared in Perspex boxes (18.5 cm x 12.5cm x 6.5cm). Boxes were then randomly assign to control or treated groups. The control group was fed ad libitum with 50% v/v apiary solution for six days whereas the treated group was fed ad libitum with a 10 ppb imidacloprid (SIGMA-ALDRICH) 50% v/v apiary solution. After six days of exposure the bees were anesthetized on ice at 4◦C. The brains were dissected in phosphate buffered saline (PBS) and immediately frozen in liquid nitrogen and stored at -80◦C. Their ovaries were checked for development to ensure that only non-reproductive workers were used [161, 175].

2.2.3 BS-seq

2.2.3.1 Genomic DNA extraction, sequencing and mapping

Genomic DNA was extracted, using QIAGEN QIAamp DNA Micro Kit following the manufacturer’s instruction, from 12 non-reproductive workers brains per biological sample. The concentration of genomic DNA was measured using a Qubit® dsDNA Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 22

BR Assay Kit (ThermoFisher Scientiﬁc, USA) and Nanodrop. Six libraries (three control and three treated samples, one from each of the three colonies, total of 72 bees) were prepared and whole-genome bisulﬁte (WGBS) sequenced on a HiSeq 2000 machine (Illumina, Inc.) at the Beijing Genomics Institute (BGI), generating 100-bp paired-end reads.

Base calling and quality scoring of the raw reads were visualized using fastQC v0.11.2 [176] and adapters were trimmed using cutadapt v1.11 [177] and Trimmomatic v0.36 [178] PE, ILLUMINACLIP:TruSeq3.fa:2:30:10, LEADING:3, TRAILING:3, SLIDINGWIN- DOW:4:15, MINLEN:36. The bisulfite treated reads were aligned and mapped to the reference genome of the bumble bee (Bter_1.0 genome - Refseq accession no. GCF_000214255.1) [179] using bismark v0.18.1 with default parameters “bismark ref genomes -1 fq1.gz -2 fq2.gz”[180] and bowtie2 v2.2.3 [181]. The duplicated reads were removed using the bismark deduplicate function; deduplicate-bismark – bam sample.bam . The methylation levels per C context (CpGs, CHHs, CHGs) were extracted using bismark’s methylation extractor function; bismark-methylation-extractor –bedGraph –gzip deduplicated.bam. Bases with low coverage (below 10 reads) in each sample were discarded. The reports files from bismark and the B. terrestris annotation file (GCF_000214255.1) were combined using the sqldf v0.4-11 library [182] in R v3.4.0 [183] to generate the distribution of methylated Cs over genomic features. For each cytosine the proportion of methylation reads over total reads was calculated.

2.2.3.2 Methylation differences between treatments

Differences in methylation between treatments were calculated using a logistic regression implemented in methylKit v1.6.1 [184] in R v3.4.0 [183]. Signiﬁcant difference in methylation was deﬁned as a site with a p value 0.01 and either a minimum of < 10% (CpGs) or 5% (CHHs, CHGs) difference between control and imidacloprid samples. The distribution of these differential methylation sites over genomic features was analysed using GenomicFeature v1.32.0 [185] and Rsamtool v1.32.0 [186]. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 23

2.2.3.3 Methylation differences in exon level

To identify exons’ methylation biases, we used the R library DEXSeq [187] and the bismark bam files. DEXSeq was designed to find differential exon usage using RNA-seq exon counts between samples. I modified the input files to methylation reads counts. Per-exon dispersions was calculated as described in Love et al. [188]. A generalized linear model was fitted for each gene and the deviances from the fit was tested using a chi2 distribution, providing a p value. This identified genes showing exons methylation biases at p 0.0001. <

2.2.4 RNA-seq

2.2.4.1 RNA extraction and Illumina sequencing

Three non-reproductive worker brains were pooled for each sample. Total RNA was isolated from 18 samples (3 colonies x 2 conditions (neonicotinoid imidacloprid and control samples) x 3 technical replicates, total 54 bees) utilizing a GenElute Mam- malian Total RNA Miniprep kit (Sigma-Aldrich) following the manufacturers’ protocol. DNA and RNAase activity was eliminated using the DNase I mini kit (SIGMA) following the manufacturer’s instruction. RNA concentration and integrity were determined by Bioanalyzer using the RNA Nano Kit (Agilent Technologies). From each sample we isolated an average of 0.8 mg of RNA. Two samples appeared degraded (shift in RNA size and multiple small fragments) and were not used in the rest of the analysis. In total, sixteen paired-end libraries (nine control and seven treated samples) were prepared and sequenced on HiSeq 2000 machine (Illumina, Inc.) at BGI Tech Solution Co., Ltd. (Hong Kong) and 100-bp paired-end reads were generated. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 24

2.2.4.2 Alignment and assemble transcripts

BGI removed adaptor sequences, contamination and low-quality reads from raw data. Base calling and quality scoring of the raw reads were visualized using fastQC v 0.11.2 [176]. The clean reads for each sample were aligned to the reference genome of bumble bee (Bter_1.0 genome - Refseq accession no. GCF_000214255.1) [179]) using Hisat2 v2.0.4 [189] with default parameters. The output .sam file was sorted and converted to a bam file using samtools v1.3.2 [190]. Aligned reads were assembled and quantified using the assembler stringtie v1.3.3b [191].

2.2.4.3 Differential gene expression analysis

A table of raw counts was generated using a Python script (https://ccb.jhu.edu/ software/stringtie/dl/prepDE.py) and analysed using DESeq2 [188] in R v3.4.0 [183] to estimate differentially expressed genes. Genes with less than 10 reads were discarded from analysis. The normalized read counts were log2 transformed. The quality of replicates was assessed by plotting read counts of samples against one another and assessing the dispersion and presence of any artefacts between samples [192]. A principal-component analysis was performed to visualize diversity between samples within treatment and between condition. The signiﬁcance level was set to an FDR value less than 0.05 using a Benjamini-Hochberg (BH) correction test [193].

2.2.4.4 Alternative splicing events

The Ballgown package v2.12.0 [194] in R v3.4.0 [183] to calculate the differentially expressed transcripts between control and imidacloprid samples. The assembled and quantiﬁed transcripts for each sample was computed in the Hisat2-StringTie pipeline using Hisat2 v2.0.4 [189] and stringtie v1.3.3b [191]. Genes with less than 10 transcripts were discarded from analysis. texpr(bg, ’FPKM’) function was performed to extract expression values as fragments per kilobase of transcript per million for transcripts. The Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 25 default statistical test in ballgown was implemented by a parametric F-test comparing nested linear models between control and imidacloprid samples and q value 0.05 < after FDR correction [195].

2.2.5 GO term enrichment and Kegg analysis

A list of GO terms for the buff tailed bumble bee were made by annotating the transcriptome using trinotate (default settings) [196] and blast2GO (against RefSeq) [197]. These lists were combined, using the pipeline implemented in Amar et al. [169] with a K value of 1. The GO enrichment was performed against the whole transcriptome bumble bee as background gene lists rather than from the transcriptome made from the control. A hypergeometric test was applied on genes differentially methylated, alternative spliced and expressed and signiﬁcant GO terms identiﬁed after Benjamini and Hochberg correction (p value 0.05) [193] using GOstats [199]. For the visualisation of < the enriched GO term I used REVIGO [200] selecting the whole UniPro database and SimRel semantic similarity measure.

The clusterprofiler v3.8.1 [201] R v3.4.0 [183] identified genes from kegg pathways using the whole UniPro database. A hypogeometric test was applied and significant kegg pathways were identified after Benjamini and Hochberg correction (q value 0.05) < [193]. 2.3 Results

2.3.1 Methylation

2.3.1.1 Sequencing, alignment and Methylation analysis

In each replicate set (3 biological replicate * 2 condition (imidacloprid and control), I obtained on average 3.4 mg of DNA. Six libraries of Whole-Genome Bisulﬁte Sequenc- ing were generated from treated samples and untreated samples. On average in each sample, we obtained 81 +/- 6 (mean +/- standard deviation) million paired end reads and each reads consists of 100 clean base pair (bp). Percentage of GC contain ranged from 22.70 to 25.17 %. The error rates of quality scores across all bases were less than 0.1% and the N (low quality) content across all bases was zero.

Bismark produced a run report which contained information about; percentage of sequences analysed at unique alignment, no alignment, multiple alignment, mapping efﬁciency, number of cytosines analyses, number of methylated and unmethylated cytosines, and percentage methylation of cytosines in CpG and non CpG sites (ﬁgure 2.1). The amount of aligned sequenced after deduplication ranged from 18.6 to 21.8 Mbp per sample with an average of 15.87% +/-2.44% (average +/-) of deduplicated reads removed. The overall sequence alignment rate, ranged between 65.67% to 69.42% with an average of 67.21 1.53%. The level of methylation calculated by Bismark [180] was ± 0.53% +/-0.05% for CpGs, 0.37% +/-0.05% for CHGs, 0.38% +/-0.07% for CHHs and 0.4% 0.06% for CNs or CHNs ((H = A, C, or T) (table 2.1). These methylation levels are ± comparable with other studies in the Hymenoptera [154, 156–158].

26 Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 27

Sample name Control 1 Control 2 Control 3 Imidacloprid 1 Imidacloprid 2 Imidacloprid 3 Mapping efﬁciency 60.8% 58.9% 62.3% 59.1% 64.1% 59.6% Overall alignment rate 65.67 67.2 67.6 66.4 65.8 69.42 Number of C’s 970811437 945459739 970895339 970895339 944435525 1007802368 % methylated CpG 0.5% 0.5% 0.6% 0.5% 0.6% 0.5% % methylated CHG 0.3% 0.3% 0.4% 0.4% 0.4% 0.4% % methylated CHH 0.3% 0.4% 0.5% 0.4% 0.4% 0.4% % methylated C’s in CpG 1683257 1566692 2005347 1760906 2023050 1702792 Methylate C’s in CHG 504447 504549 655267 587376 651443 601883 Methylated C’s in CHH 1677551 1632438 2166723 1927590 2259661 1880473 Unmethylated C’s in CpG 329533837 330101882 332407043 329591554 332191620 354353309 Unmethylated C’s in CHG 154875738 150999413 155523530 151585852 160207810 162529206 Unmethylated C’s in CHH 482536607 460654765 478137429 458982247 510468784 457021788

TABLE 2.1: Shows number of sequences analysed, number of sequences with a unique best alignment, statistics summarising the bisulﬁte strand the unique best alignments, number Cs and percentage methylation of cytosines in CpG, CHG or CHH context (where H can be either A, T or C).

The distribution of CpG methylation shows a mild bimodal distribution with the vast majority of sites being not or only modestly methylated and a few fully methylated (figure 2.2 A). Contrary to CpGs, most CHHs and CHGs sites are only modestly methylated (figure 2.2 B,C). Methylated CpGs are more abundant in coding regions (seven fold) and exons (five fold) than introns (figure 2.2 A). This distribution pattern is not repli- cated in CHHs and CHGs where the proportion of methylated Cs is equally distributed amongst different genetic features (figure 2.2 B,C). Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 28

FIGURE 2.1: Examples of histogram of percentage methylation per cytosine. His- togram of % methylation per CpG (A), CHH (B) and CHG (C) for a control sample (C1). CpG % methylation shows a mild bimodal distribution with most sites modestly methylated and few fully methylated. Contrary to CpG, most CHH and CHG are modestly methylated. Both forward and reverse reads are reported.

2.3.1.2 Methylation differences between treatments

I identiﬁed 79 , 86 and 16 genes deferentially methylated in genomic contexts CpGs, CHHs and CHGs respectively (p value 0.01 and percentage difference of a minimum < 10% for CpGs and 5% for CHHs and CHGs) due to treatment. Larger differences of methylation were found for CpGs (38.99% difference) than CHHs (12.77% difference) and CHGs (11.76% difference). The proportion of differentially methylated CpGs is only marginally higher for genes (94.94%) than exons (86.08%) suggesting that almost all differentially methylated CpGs are in exons. Conversely, for CHHs and CHGs the proportion of differentially methylated Cs in genes is twice (CHG genes:15.92%, exons:7.35%) and three times (CHH genes:62.79%, exons:17.44%) of that in exons (ﬁgure 2.3). Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 29

A B

Control Neonicotinoid

FIGURE 2.2: Methylated Cs distribution. Average proportion of methylation reads SD per CpG (A), CHH (B) and CHG (C) positions over genomic features. Control ± samples in black and neonicotinoid treated samples in grey.

Several genes have differentially methylated Cs in different genomic contexts (CpG, CHH, CHG), see (table 2.2). mig-15, Robo2, ten-m, Sb and MXD1 for example are genes in both CpG and CHH lists. Interestingly, mig-15, Robo2 and ten-m are all involved in axon/neuron formation [202, 203]. MXD1 is an inhibitor of Myc and was found to be interacting with rDNA Myc binding regions [204]. Recently, the expression of MXD1 gene was found to be controlled by SIRT1 and DNMT3B [205]. Sb is involved in the RhoA signaling pathway in imaginal discs. Type II transmembrane serine proteases like Sb are necessary for normal intracellular signaling during development [206].

Some genes have multiple differentially methylated Cs: lachesin, mig-15, ten-m, neur, STAT C-like and the unknown genes LOC100650751, LOC100646518 , LOC105667086 all have multiple differentially methylated CHHs; UBALD2, Sb, KAZN, ets DNA-binding protein pokkuri, LOC100649677, LOC100646399 , LOC110120039 have multiple differentially methylated CpGs. lachesin is involved in mushroom body and antennal lobe Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 30

FIGURE 2.3: Shows proportion of differentially methylated Cs over genomic features for CpG (orange bar charts), CHH (red bar charts), CHG (grey bar charts). a CHG and CHH with p value 0.01 were used in the chart, regardles of the % of difference be- < tween neonicotinoid and control samples development in Locusta migratoria [207] and neur is part of Notch signaling [208]. Kazrin (KAZN) is involved in the regulation of RhoA in Xenopus [209] and pokkuri has been associated with retina development in Drosophila [210].

Other genes in these lists are particularly interesting for their roles in neuron-neuron communication (Dopamine receptor 2: Drd2 and the circadian clock (PDF receptor: Pdfr [211]). Decapentaplegic (dpp) is involved in tissue growth [212] and Notch1 (neurogenic locus notch homolog protein 1) is important for synaptic plasticity and for olfactory behaviour [213, 214]. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 31

Gene CpG CHH CHG Note mig-15 Y Y Robo2 Y Y ten-m Y Y Sb Y Y MXD1 Y Y cadherin-87A Y Y * KAZN Y Y midasin Y Y lachesin Y Y Y * lachesin, transcript variant X2 Y Y * Rbp6 Y Y hdc Y Y Dh44 Y Y ZNF800 Y Y * FAR1 Y Y * neur Y Y * uncharacterized LOC105667086 Y Y * uncharacterized LOC105667080 Y Y * uncharacterizedLOC105667085 Y Y Y * uncharacterized LOC100646399 Y Y Y uncharacterized LOC105666194 Y Y uncharacterized LOC105666996 Y Y * uncharacterized LOC110120048 Y Y uncharacterized LOC100649677 Y Y

TABLE 2.2: Genes with differentially methylated Cs (p value less than 0.01) in multiple contests (CpG, CHH, CHG). * less than 5% differentially methylated CHG, p value less than 0.01

2.3.2 Methylation differences in exon level

I also asked whether DNA methylation was associated with exon usage. To answer this question I used the R library DEXSeq [187]. This library is designed to identify differential usage of exons expression (alternative splicing) using RNA-seq reads. I modified the methylation bismark deduplicated bam files (with reads per each methylated Cs) and the gff file with samtools [190], python and cufflinks [215] to serve as input for this library. The library adjusts for coverage biases and perform a dispersion correction and fits a GLM using as variables exon and condition. The methylation fold change per exons are calculated based on the coefficient of the GLM fit. I identified 5 genes with p Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 32 value 0.0001 ( figure 2.4). These genes are facilitated trehalose transporter Tret1-2 < homolog (Tret1-2) that is involved in the sugar methabolism in the brain of fruit fly [216], CREG1 important for cellular differentiation and fly development [217], beclin 1-associated autophagy-related key regulator (Atg14) involved in autophagy [218], me- diator of RNA polymerase II transcription subunit 15 (Med15) that mediate the transcription of DPP targets [219] and protein phosphatase Slingshot (Ssh) that control actin polymerization. CREG1 shows a methylation “switching” for Exon(E)002, E003 and change in level of methylation for E001. Similar pattern can be see for the other 4 genes. In particular Atg14 E002, E003, Med15 E001, E003, E007, E009 , Ssh E004, E008 showed methylation “switching”. 33

Methylation fold change Methylation fold change xnuaefl hne r acltdbsdo h ofcet faGMfi.Frec eeteito xnantto o each for annotation exon intron the gene each For fit. GLM a of coefficients the on based calculated are changes fold usage Exon F IGURE .:MtyainEossicig ehlto odcag o oto rd n e bu)smlspreosi v genes. five in exons per samples (blue) Neo and (red) control for change fold Methylation switching. Exons Methylation 2.4: Tret1 CREG1 - 2

Methylation fold change Methylation fold change rncit sshown. is transcripts Atg14 Med15

Methylation fold change Ssh Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 34

2.3.3 GO term and KEGG pathway analysis

From the differentially methylated genes, I found 147 (CpGs), 215 (CHHs), 54 (CHGs) enriched gene ontology terms (after Benjamini and Hochberg corrected test, P value < 0.05) associated with the various differentially methylated Cs.

In biological processes; for CpGs, the most represented categories were “somatic stem cell division”, “developmental process”, “biological adhesion”, “locomotion”, “endocytosis” and “primede novoprime NAD biosynthesis from aspartate” (figure 2.6). For CHHs, the most represented categories were “karyosome formation”, “molybdate ion transport”, “lactose biosynthesis”, “negative regulation of locomotion”, “cell-cell adhesion”, “ntibiotic metabolism”, “alternative mRNA splicing, via spliceosome” and “developmental processes” (figure 2.7). The biological processes for CHGs were “regulation of synaptic vesicle fusion to presynaptic membrane”, “regulation of synaptic vesicle fusion to presynaptic membrane”, “homophilic cell adhesion via plasma membrane adhesion molecules”, “DNA damage induced protein phosphorylatio” and “histone mRNA catabolism” (figure 2.8).

I have found 21 signiﬁcant (hypogeometiric probability p 0.0001) overlapping GO < terms between the CpGs and CHHs lists, 5 overlaps between CpGs and CHGs (p < 0.0001) and 8 overlaps between CHHs and CHGs (p 0.0001). The overlapping GO < terms were mostly associated with “cell morphogenesis involved in neuron differentiation” (CpGs-CHHs), “stem cell fate commitment” (CpGs-CHGs), “terminal branching” and “cytoskeletal matrix organization at active zone” (CHHs-CHGs).

The number of enriched GO terms were smaller for molecular function: 33 (CpGs), 49 (CHHs) and 5 (CHGs). The most represented categories were “phosphatidylinositol 3-kinase regulatory subunit binding”, “alpha2-adrenergic receptor activity”, “phosphatidylinositol 3-kinase regulatory subunit binding”. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 35

In molecular function For CpGs, the most represented categories were; “ phosphatidylinositol 3kinase regulatory subunit binding”, “alpha2adrenergic receptor activity”, “nicotinamidenucleotide adenylyltransferase activity”, “RNA polymerase II regulatory region sequencespecific DNA binding”, “cGMP binding”, “protein binding”, “ribosome binding”, “Ltyrosine transmembrane transporter activity”, “ transcription factor activity, RNA ”polymerase II distal enhancer sequencespecific binding and “neuropeptide binding” (figure 2.9). Most enriched GO terms in molecular function for CHHs were; “eceptor activity”, “ alphaketoglutarate transmembrane transporter activity”, “ mannokinase activity”, “longchainenoylCoA hydratase activity”, “Crich sin- glestranded DNA binding”, “ longchain3hydroxyacylCoA dehydrogenase activity” and “dipeptidylpeptidase activity” (figure 2.10). For CHGs, the most GO terms were; “protein serine/threonine kinase activity”, “voltagegated calcium channel activity”, “actin monomer binding”, “glucosidase activity”, “ dopamine betamonoooxygenase activity”, “poly(U) RNA binding” and “photoreceptor activity” (figure 2.11). I have found very little overlaps between molecular function GO terms: 3 overlaps between CpGs and CHHs (p value<0.0001; “MAP kinase activity”, “adrenergicreceptor activity” and “filamin binding”).

No KEGG pathways were over represented for differentially methylated CHGs or CHHs (q value 0.05). A mitogen-activated protein kinase (MAPK) signaling cascade (PATH- < WAY: bter04013) was over represented in differentially methylated CpGs at q value = 0.008 (ﬁgure 2.5). Four genes were over-representative with this KEGG pathway including; protein decapentaplegic (LOC100646366), serine/threonine-protein kinase mig- 15 (LOC100650469), ets DNA-binding protein pokkuri (LOC100651145) and uncharacterized LOC100644466. In fruit ﬂies, two of the three MAPK pathways (JNK and p38) are activated by environmental stresses [220]. The MAPK signaling cascades plays a crucial role in regulation of gene expression and typically are linked to many important biological function, for examples cell shape change, longevity, immune respond, egg polanty, bacterial infection, heat shock and eye development. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 36

FIGURE 2.5: A KEGG pathway diagram shows the relationships of genes differentially methylated and map pathway of mitogen-activated protein kinase (MAPK) signaling cascade pathway : bter04013. A green rectangles show functional genes common of gene product (an enzyme), Red outline shows pseudogene formation and white rectangle shows absence gene. REVIGO Gene Ontology treemap

regulation 'de novo' NAD behavioral anterograde hematopoietic spindle stomatogastric synaptic growth aspartate epithelium migration of synapse biosynthetic process at neuromuscular response trans−synaptic stem cell pole body nervous system peptidyl−serine structure or from aspartate metabolic somatic stem cell division to ethanol signaling homeostasis organization development dephosphorylation junction activity process primede novoprime NAD biosynthesis from aspartate

ATP generation from negative formation of poly−ADP−D−ribose chemotaxis response to response to tube single organism regulation of NAD biosynthetic process cell morphogenesis anatomical wounding corticosteroid development signaling fusion cell fate involved in neuron boundary positive regulation of differentiation specification cell differentiation establishment homophilic cell of synaptic adhesion via plasma stem cell fate endocytosis branched duct positive regulation regulation of synaptic vesicle vesicle membrane adhesion commitment Roundabout epithelial cell fate of regulation of establishment of membrane localization molecules circadian signaling pathway determination, open phosphatidylinositol vasoconstriction temperature tracheal system 3−kinase signaling cell polarity organization homophilic cell homeostasis negative regulation of adhesion via plasma endocytosis olfactory bulb SRP−dependent neuroblast proliferation somatic stem cell division membrane adhesion interneuron sex comb plasma negative negative regulation negative vesicle−mediated cotranslational molecules response to of compound eye 37 transport in protein targeting development development membrane regulation retinal cell regulation of alcohol synapse to membrane, tubulation of behavior programmed cell death GTPase activity negative regulation of translocation negative chemotaxis regulation of immature salivary gland R8 cell−mediated cellular regulation of positive regulation of clathrin−dependent T cell proliferation boundary specification photoreceptor animal organ response to ARF protein response to copper ion adenylate cyclase activity endocytosis organization formation dexamethasone involved in G−protein coupled signal receptor signaling pathway magnesium ion homeostasis positive stimulus transduction regulation of protein epithelial cell positive multicellular regulation of maintenance of developmental cell negative regulation organismal movement of cell or glycosylation compound eye proliferation regulation of process adaxial/abaxial cell projection presynaptic of transcription proliferation subcellular component in Golgi corneal lens involved in multicellular process pattern formation organization active zone from RNA polymerase development Malpighian tubule II promoter organismal structure morphogenesis process

establishment toll−like receptor specification of determination rostrocaudal cell chaeta oenocyte cytoskeleton or maintenance biological adhesion locomotion epithelium development 7 signaling pathway segmental identity, of genital disc neural tube communication development development organization of cell labial segment primordium patterning polarity

FIGURE 2.6: GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing < signiﬁcant differentially methylated genes at CpG sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. REVIGO Gene Ontology treemap

positive regulation retrograde Wnt signaling positive of Wnt signaling regulation of trans−synaptic pathway involved in lactose L−serine regulation of neuronal signal body regulation positive pathway involved in presynapse signaling by midbrain molybdate ion biosynthetic biosynthetic karyosome formation circadian of protein regulation of transduction dopaminergic neuron morphogenesis transport sleep/wake cycle dorsal/ventral axis assembly trans−synaptic localization ruffle assembly process process differentiation specification protein complex to synapse positive positive regulation mannose disaccharide NMDA glutamate regulation of primitive positive primitive streak glucose regulation of metabolic metabolic presynaptic terminal button of oxidative regulation of receptor presynaptic erythrocyte transmembrane receptor−mediated instar larval active zone stress−induced formation process process organization clustering membrane differentiation synaptic vesicle endocytosis or pupal neuron death transport organization organization clustering molybdate ion transport leukotrienelactose biosynthesisfructose morphogenesis regulation of receptor−mediated positive metabolic metabolic adenylate trachea cellular embryonic retina external presynaptic endocytosis nucleoside regulation of cyclase−activating process process cartilage calcium ion morphogenesis in genitalia cytosolic transmembrane age−dependent dopamine receptor involved in rRNA processing startle response camera−type eye morphogenesis homeostasis signaling pathway morphogenesis calcium ion cholesterol transport L−ascorbic acid response to lysine catabolic concentration transport manganese ion metabolic metamorphosis oxidative stress process transport process beta−catenin regulation of positive cellular lactam dopaminergic excitatory regulation of tissue innate immune destruction respiratory N−methyl−D−aspartate regulation of homocysteine oxoacid canonical Wnt catabolic neuron synapse selective glutamate neurotransmitter synaptic vesicle morphogenesis response complex gaseous metabolic metabolic signaling pathway process differentiation assembly receptor activity transport budding disassembly exchange process process involved in karyosome formation terminal neural crest cell 38 negative cytoskeletal synaptic negative regulation branching, open differentiation Roundabout mating regulation of clathrin coat matrix vesicle negative regulation of MyD88−dependent tracheal system exocrine system cellular response response to positive multicellular signaling behavior, sex assembly organization at membrane to cholesterol odorant of locomotion toll−like receptor regulation of canonical Wnt development discrimination organismal pathway active zone organization signaling pathway heterotypic signaling pathway process cell−cell involved in female adhesion anterograde imaginal developmental postsynaptic negative regulation of regulation of cell germline neural crest response to trans−synaptic appendage specialization regulation membrane cell−cell proliferation disc−derived head growth ring canal formation response to development Thyroglobulin negative regulation of locomotion signaling appendage development involved in organization catecholamine of negative hyperpolarization adhesion formation triiodothyronine morphogenesis morphogenesis chemotaxis cellular response to cholesterol negative dorsal/ventral single synaptic negative cell−cell pericardium regulation cellular cellular response negative salivary gland axon guidance anatomical aging organism regulation genitalia gland morphogenesis target adhesion sleep of response hyperosmotic to nitric oxide regulation of of hemocyte morphogenesis structure signaling differentiation development development to stimulus smooth muscle cell attraction regression response response to apoptotic process negative cell surface protein compound regulation chemical protein kinase synaptic cellular alternative telencephalon regulation of axis elongation eye corneal of Notch C−activating receptor complex post−embryonic G−protein coupled developmental mRNA splicing, antibiotic male germ cell involved in vesicle response to development subunit development lens signaling receptor signaling response to via signaling pathway targeting process metabolism proliferation somitogenesis development interleukin−4 pathway organization pathway insecticide spliceosome

FIGURE 2.7: GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing < signiﬁcant differentially methylated genes at CHHs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. REVIGO Gene Ontology treemap

regulation of synaptic regulation of cell fate commitment positive regulation muscle cell fate ventral cord vesicle fusion to auxin polar transport asymmetric sensory organ precursor involved in pattern epithelium development of endocytosis specification development presynaptic membrane cell division cell fate determination specification homophilic cell adhesion via plasma membrane adhesion molecules

homophilic cell adhesion via plasma membrane regulation of maintenance of positive regulation regulation of alpha−amino−3−hydroxy−5−methyl−4−isoxazole ribosomal protein pericardial nephrocyte negative regulation adhesion propionate selective presynaptic active of integrin−mediated compound eye molecules glutamate receptor activity import into nucleus differentiation gland development of hemocyte zone structure signaling pathway photoreceptor sensory organ precursor cell fate determinationdifferentiation hepatocyte differentiation development

cell adhesion regulation of synaptic vesicle fusion to presynaptic membrane involved in peripheral nervous 39 negative regulation epithelial cell type heart regulation of rubidium Rab protein signal protein negative of actin filament system development specification, open endosome to morphogenesis ion transport transduction homooligomerization regulation of depolymerization tracheal system melanosome epithelial transport cell migration

pre−B cell allelic exclusion Malpighian tubule tip regulation of regulation of regulation of Notch histone cell differentiation neuron remodeling platelet degranulation cell division signaling pathway deubiquitination post−embryonic development histone mRNA regulation of cellular response catabolism sulfate DNA damage induced protein vascular endothelial to vascular protein−chromophore cytoskeletal matrix neuronal stem DNA damage induced protein phosphorylation transmembrane ruffle organization protein phosphorylation autophosphorylation growth factor endothelial growth linkage organization at active zone cell division transport signaling pathway factor stimulus

FIGURE 2.8: GO term enrichment for biological Processes (BP). Enriched BP for GO terms (p 0.05) associated with genes containing < signiﬁcant differentially methylated genes at CHGs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. REVIGO Gene Ontology treemap

phosphatidylinositol 3−kinase nicotinamide−nucleotide nicotinate−nucleotide regulatory subunit binding identical protein binding SH3 domain binding adenylyltransferase activity adenylyltransferase activity L−tyrosine transmembrane protein binding transporter activity

phosphatidylinositol 3−kinase regulatory subunit binding

nicotinamide−nucleotide adenylyltransferase activity

protein kinase A TPR domain binding regulatory filamin binding

subunit binding polypeptide phosphatidylinositol−4,5−bisphosphate lysophosphatidic acid N−acetylgalactosaminyltransferase 3−kinase activity acyltransferase activity activity

neuropeptide binding ribosome binding 40

alpha2−adrenergic receptor activity intracellular cGMP MAP kinase kinase ARF guanyl−nucleotide activated cation RNA polymerase II regulatory region kinase kinase activity exchange factor activity transcription factor activity, RNA channel activity sequence−specific DNA binding polymerase II distal enhancer alpha2−adrenergic receptor activity sequence−specific binding RNA polymerase II regulatory region ARF guanyl−nucleotide sequence−specific DNA binding exchange factor activity cGMP binding

ubiquitin−protein calcitonin receptor activity transferase phosphatidylinositol−3,5−bisphosphate 7S RNA binding cyclic nucleotide−gated ion channel activity activator activity 3−phosphatase activity

FIGURE 2.9: GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes < containing signiﬁcant differentially methylated genes at CpG sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. REVIGO Gene Ontology treemap

dopamine neurotransmitter neurotransmitter inositol 1,4,5 1−phosphatidylinositol receptor activity, receptor activity apolipoprotein clathrin binding trisphosphate binding binding receptor activity coupled via Gs binding phosphoglycerate long−chain−3−hydroxyacyl−CoA dehydrogenase activity dehydrogenase long−chain−3−hydroxyacyl−CoAactivity inositol 1,4,5 trisphosphate binding dehydrogenase activity clathrin binding low−density adrenergic coreceptor lipoprotein filamin binding receptor activity receptor activity activity glucose binding NAD binding coreceptor activity involved in receptor activity receptor canonical Wnt signaling pathway L−malate dehydrogenase activity binding

protein dimerization activity

transmembrane receptor protein Wnt−activated neuropeptide Y tyrosine C−rich 41 peptidyl−dipeptidase hyaluronic tyramine receptor activity receptor activity receptor activity dipeptidyl−peptidase activity single−stranded phosphatase activity acid binding activity DNA binding dipeptidyl−peptidase activity MAP kinase kinase mannokinase activity kinase kinase activity

alpha−ketoglutarate molybdate ion transmembrane xenobiotic transmembrane metallodipeptidase activity transporter transporter activity transporter activity mannokinase activity activity acetyl−CoA signal transducer activity C−acyltransferase alpha−ketoglutarate transmembrane transporter activity enzyme activity protein inhibitor tyrosine long−chain−enoyl−CoA enoyl−CoA activity kinase long−chain−enoyl−CoA hydratase activity D−glucose transmembrane transcriptional repressor dehydroascorbic acid activity hydratase activity hydratase activity store−operated calcium activity, RNA polymerase II transporter activity transporter activity channel activity diacylglycerol kinase activity transcription regulatory region sequence−specific binding

FIGURE 2.10: GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes < containing signiﬁcant differentially methylated genes at CHHs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. REVIGO Gene Ontology treemap

1−phosphatidylinositol−3−kinase activity voltage−gated calcium channel activity N−1−naphthylphthalamic acid binding

calcium−dependent protein voltage−gated calcium channel activity calcium−dependent protein serine/threonine kinase activity serine/threonine kinase activity

secondary active sulfate transmembrane transporter activity calmodulin−dependent protein kinase activity

42 dopamine beta−monooxygenase poly(U) RNA binding activity

maltose alpha−glucosidase activity

glucosidase activity actin monomer binding actin monomer binding Notch binding

glucosidase activity photoreceptor activity

FIGURE 2.11: GO term enrichment for Molecular function. Enriched molecular function for GO terms (p 0.05) associated with genes < containing signiﬁcant differentially methylated genes at CHGs sites. These rectangles are joined into different coloured ‘superclus-ters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 43

2.3.4 Differential expression analysis

RNA extraction, sequencing and mapping efﬁciency

In each replicate set (3 biological replicate * 2 condition (neonicotinoid imidacloprid and control) * 3 technical replicate, I obtained on average 0.8 mg RNA. The RIN quality for 16 samples (7 treated and 9 untreated samples) showed that the predominant 28s and 18s rRNA peaks and bands with to 2 of 28s/18s ratio in electropherogram and gel images. While 2 samples in the treated group showed a shift in the RNA size and degraded with small fragments of RNA. As regards RNA Integrity Number (RIN), all 16 samples were zero while 2 samples in treated group were 5.4 and 6.4. In honey bees and most insects, 28S rRNA consists of two separate fragments that are hydrogen-bonded together [221]. Sixteen libraries (7 treated samples and 9 untreated samples) at high quality of whole transcriptome shotgun sequencing were obtained.

In each samples, I obtained 21 +/- 7 million clean 100 bp reads. FASTQC visualization for all samples showed, probability of incorrect base call under quality score (Q30%) was 1 in 1000 and percentage of GC content ranged between 37.9 to 40 %. The error rates of quality scores across all bases were less than 0.1% and N (low quality) contents across all bases was zero. These quality checks suggested that the sequences were of high quality. We used the Hisat2 aligner to align raw reads to the bumble bee reference genome. Alignment rate was 93.6% (92.1 to 94.1) on average, with 88.9% reads aligned concordantly exactly 1 time.

The only exception is the control sample 7 (untr-7) that clustered with the imidacloprid treated samples. I have fixed this differences in expression by removing differences (artefacts) between sample 7 (untrt-7) with rest of untreated samples [192]. With this technique I I identified genes that differ between replicates in same condition which the difference may due to sequencing technical error. As shown in figure below ( 2.13), there are few artefact genes between samples in same condition. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 44

FIGURE 2.12: PCA plots showing correlation between samples and distance between neonicotinoid and control condition. Red dots representing imidacloprid samples and blue points representing control samples

I considered these genes may effect differential expression analysis. The code available in appendix (A.1) was used to remove outlier genes between samples. The code was provided and generated to MIBTP students by Sascha Otts group (they ran the RNASeq on the MIBTP). Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 45

FIGURE 2.13: Scatter plot showing log2 counts for each gene between sample 7 and 6 in control samples. The black line represents the real relationship between the x and values. The red line shows how extreme a value would need to be considered as an artefact.

2.3.4.1 Differential gene expression

Low count genes ( 10 reads) were ﬁltered resulting in a list of 10,772 genes. Both the < euclidian distance and PCA analysis showed that the samples segregated by treatment (imidacloprid vs control (ﬁgure 2.12).

A negative binomial generalized linear model test was implemented to find differentially expressed genes between control and imidacloprid treated samples. After FDR correction at p value 0.1, I have found 597 differentially expressed genes genes: 264 < genes up regulate and 333 down regulated. At p value 0.05, 378 genes were differ- < entially expressed genes: 216 genes up regulated and 162 down regulated (figure 2.14). Fold changes for most of differentially expressed genes was less than one (figure 2.15). However, 18 genes showed log2 fold change in gene expression higher than 1. Four of these genes belong to the homeotic box gene (Hox) family and are all down regulated. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 46

Hox genes play important role in Drosophila head development [222]. Miss expression of the homeotic gene antennapedia for example results severe alteration of the fly body plan (e.g. legs grow from the fly’s head instead of antennae). Up regulated genes in imidacloprid treated bees include apyrase that hydrolyzes ATP to AMP and ionotropic receptor 25a that is involved in circadian clock resetting in Drosophila [223]. Down regulated genes include numerous apotosis related genes including caspase-1, sushi and Anoctamin. lethal(2) essential for life (Efl21) displayed the highest down regulation. Efl21 has been found to be involved in foraging behaviour in bees [224]. Six cytochrome P450 genes showed differentially expressed, four of them showed up regulation and others down regulation. In Drosophila melanogaster an over expression of detoxification genes correlates with increased survival in resistance to pesticide, while cytochrome knock-down dramatically reduce survival [225]. Cytochrome P450s of the CYP9Q subfamily were recently shown to be responsible for bee sensitivity to neonicotinoids [226]. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 47

condition condition LOC100645813 ctrl LOC100650704 0.5 Neo LOC100646260 LOC100647255 LOC100644229 LOC100649904 0 LOC100645243 LOC100647245 LOC100651332 LOC100647296 −0.5 LOC100648603 LOC100650064 LOC100651992 LOC100649414 LOC100642931 LOC100644732 LOC100647179 LOC100645209 LOC100650431 LOC100650153 LOC100650597 LOC100650650 LOC100648691 LOC100646400 LOC100642296 LOC100631088 LOC100648391 LOC100645024 LOC100642739 LOC100645386 LOC100652167 LOC100649106 LOC100642770 LOC100645609 LOC100650672 LOC100651204 LOC100644036 LOC100646517 LOC100648756 LOC100648970 LOC100642602 LOC100647779 LOC100648549 LOC100644995 LOC100642544 LOC100651427 LOC100649077 LOC100651530 LOC100642779 LOC100650142 LOC110120070 LOC100650309 LOC100645983 LOC100648157 LOC110119287 LOC105666599 LOC100642474 LOC100651433 LOC100643972 LOC100646094 neo_4 neo_6 neo_7 neo_5 neo_3 neo_8 neo_9 untrt_1 untrt_4 untrt_5 untrt_2 untrt_8 untrt_9 untrt_7 untrt_3 untrt_6

FIGURE 2.14: Heatmap tree showing 60 genes. 30 genes that are up regulated and 30 genes are down regulated. The yellow means high level of gene expression while blue shows reduced relative gene expression. Neo; neonicotinoid and ctrl; control Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 48

FIGURE 2.15: Volcano plot showing result data with log fold change, The x axis is the fold change (in log2 scale); the y axis is p-value (in log10 scale). Each dot on the plot is a single gene. Colour coding is based on adjusted p value, black dots genes p value greater than 0.1 and red dots genes with p value less than 0.1 Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 49

2.3.4.2 Go terms and KEGG pathway

For genes (378 genes) differentily expressed, I found 209 enriched biological process GO terms, 83 enriched molecular function GO terms and 27 enriched Cellular com- ponents GO terms. For biological process GO terms, the nine most represented categories are: negative regulation of neuromuscguelnaesr synaptic transmission, energy reserve metabolism, protein activation cascade, anion transmembrane transport, cel- lullar glucan metabolism, reactive oxgyen species metabolism, single organism process, localization and generation of prescusor metabolism and energy . For molecular function GO terms, the most represented categories are: oxidoreductase activity, organic anion transmembrane transporter activity, cAMP response element binding, metallopeptidase activity, 1,4-alpha-glucan branching enzyme activity, histidine decarboxylase activity, acyl binding, CoA ligase activity, cargo receptor activity and epidermal growth factor receptor binding. For Cellular Component GO terms correlated to: juxtaparanode region of axon (sperm midpiece), cell periphery, mitochondrial inner membrane presequence translocase complex, respiratory chain, intrinsic component of membrane, integral component of plasma membrane and membrane.

Despite the many BP,MF and CC functions, I found no signiﬁcant enriched Kegg pathway (q value 0.05). < −ve_regulation_of organic −ve_regulation_of cellular sensory mRNA glutathione cellular acid cell viral response_to organonitrogen cytoplasmic perception_of splicing, metabolic response_to response_to_interferon−beta catabolic death genome protein unfolded compound translation pain via process starvation process replication activation protein oxidation−reduction metabolic spliceosome cascade process process

development_of establishment_of cellular histone creatine secondary pole cellular response_to apoptotic H3−K4 hypoxanthine prolactin cellular metabolic male plasm protein response_to laminar process demethylation, salvage signaling amide process sexual mRNA processing metal fluid female trimethyl−H3−K4−specific pathway metabolic characteristics localization ion shear germ−line process stress stem NAD cell response_to_topologically_incorrect energy multicellular multicellular biosynthesis asymmetric lipid protein reserve peptide organismal organismal via nonfunctional division A amyloid metabolic biosynthetic macromolecule protein nicotinamide rRNA response_to cellular biosynthetic response_to_insulin−like process process precursor catabolic metabolic riboside decay endoplasmic growth response_to −ve_regulation_of process protein process process salvage reticulum factor external cofactor −ve_regulation_of catabolic pathway stress stimulus stimulus metabolic transcription_from process regulation_of process RNA nuclear−transcribed male low−density secondary polymerase mRNA single−organism germ−line lipoprotein renal branching, II catabolic cellular pyridine catabolic sex process, particle tubular open response_to nucleotide promoter determination acute_inflammatory process non−stop receptor secretion tracheal cellular by response vascular metabolic catabolic system response_to mechanosensory cell histone decay endothelial process glutamate process chemical behavior chemotaxis modification growth metabolic stimulus +ve_regulation_of factor process regulation_of vascular mesendoderm specification_of stimulus oocyte generation_of wound development segmental +ve_regulation_of morphogenesis precursor healing identity, catabolic metabolites thorax process neurotransmitter neutral galactosylceramide embryo and single−organism loading_into amino metabolic implantation energy anion transport synaptic acid process angiogenesis_involved_in angiotensin mitochondrial wound polycistronic transmembrane vesicle transport catabolic protein xanthine healing mRNA transport process_in catabolic catabolic mitochondrial processing blood process process membrane organization aromatic dehydroascorbic anion amino developmental acid glutamate ADP carboxylic maturation transport acid chorionic mitochondrial small transport decarboxylation_to metabolic acid dicarboxylic transport trophoblast tRNA protein molecule succinate process metabolic acid cell 3'−end digestion metabolic process catabolic transmembrane differentiation processing process process

50 transport protein protein regulation_of transepithelial localization_to localization_to zinc L−ascorbic +ve_regulation_of +ve_regulation_of +ve_regulation_of perinuclear +ve_regulation_of ciliary ion acid membrane androgen dendritic dendritic region_of +ve_regulation_of protein −ve_regulation_of_interleukin−12 membrane transport transport phospholipid organization production receptor cell spine cytoplasm −ve_regulation_of leukocyte kinase intracellular scrambling activity chemotaxis maintenance neuromuscular differentiation B protein synaptic signaling transmembrane transmission transport regulation_of protein +ve_regulation_of +ve_regulation_of cell clustering_of endonucleolytic transcriptional glutamate−cysteine differentiation_involved_in localization_to protein homotypic voltage−gated cleavage_involved_in xenobiotic melanosome start salivary cell−cell ligase juxtaparanode localization_to auditory gland potassium tRNA monovalent_inorganic transport transport site activity region_of rhabdomere adhesion development channels processing cation selection_at lipid receptor axon transport RNA homeostasis cell differentiation −ve_regulation_of polymerase epithelial epidermal II cell growth +ve_regulation_of meiotic −ve_regulation_of promoter maturation_involved_in factor lactation nuclear pH membrane salivary receptor permeability envelope elevation cellular generation_of DNA−templated gland signaling disassembly glucan polysaccharide thyroid development precursor transcriptional pathway metabolic metabolic single−organism hormone protein +ve_regulation_of metabolites start process process process homotetramerization −ve_regulation_of and transport site MHC gamma−delta protein_insertion_into low−density energy selection class T ER lipoprotein II cell membrane vascular particle biosynthetic activation by endothelial clearance process +ve_regulation_of GPI_attachment growth sequence endothelial factor reactive lysosomal +ve_regulation_of carbohydrate cell triglyceride −ve_regulation_of gene receptor−2 glycerol oxygen lumen biosynthetic proliferation ATPase semaphorin−plexin aspartyl−tRNA conversion_of regulation_of signaling biosynthetic homeostasis process species localization acidification activity aminoacylation autophagosome signaling immunoglobulin pathway process metabolic size pathway genes process

anion transmembrane transport cellular glucan metabolism energy reserve metabolism generation of precursor metabolites and energy localization negative regulation of neuromuscular synaptic transmission protein activation cascade reactive oxygen species metabolism single−organism process

FIGURE 2.16: GO term enrichment for Biological Processes. Enriched BP for GO terms (p <0.05) associated with genes containing signiﬁcant differentially expressed genes. These rectangles are joined into different coloured ‘superclusters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 51

2.3.5 Differential splicing of isoforms

A total of 22,040 isoforms were found to be expressed from the 10,072 genes found in the sixteen samples. After ﬁltering all transcripts that them expression were less than 10 reads. 7,237 genes showed at least ten transcripts in all samples. The expression patterns of treated and untreated samples were more similar to each other after performing log transformation for fpkm. While some genes in sample number 10 in control group were less number of expression compare with other samples (ﬁgure 2.17).

FIGURE 2.17: Distribution of FPKM values across imidacloprid and control. samples from the same condition are shown in the same color: imidacloprid samples in blue, and control samples in orange.

I have found 25 genes, differentially alternatively spliced between control and imidacloprid samples at qvalue 0.05 after FDR correction. Most of these genes shown < Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 52 less than 2 fold changes differential expression in comparing between treated and untreated samples. While gene LOC100642739 showed more than 2 fold changes in expression in bees exposed to neonicotinoid imdacloprid compare with bees were fed on shame sugar water (figure 2.18 A). It is trehalose hydrolysis enzyme, called facilitated trehalose transporter Tret1-2 homolog gene. This enzyme exists as sugar component of haemolymph in most insects and it has been identified to serve as a mobile energy source for flight in fruit fly [227]. Many studies investigated the advantage of trehalose in protecting organisms against several environmental stresses in particular, oxidation, anoxia, and desiccation [228], neonicotinoid pesticides [124]. Cytochrome b5 domain-containing protein 2 homolog showed four different isoform (figure 2.18 B). Again these genes capable of performing oxidation and reduction [226]. It has been widely documented that reductase of these genes increase resistance to oxidative stress and life span in Drosophila melanogaster [229]. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 53

A Control Neonicotinoid

B Control Neonicotinoid

FIGURE 2.18: These plots show levels transcript structures of gene proﬁle between neonicotinod and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow. A; LOC100642734 gene (facilitated trehalose transporter Tret1-2 homolog) and B; LOC100651516 gene neuferricin or Cy- tochrome b5 domain-containing protein 2 homolog

The most striking result to emerge from the data is that LOC100646094 (gene; peptide methionine (S) -S-oxide reductase activity) showed about 2 fold change expression in the workers of bumble bees which exposed to neonicotinoid imidacloprid compare with bees which were fed on shame sugar water (ﬁgure 2.19). The main function of this gene repair oxidative damage to proteins that have been inactivated by oxidation stress [230, 231]. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 54

transcripts from gene (LOC100646094 peptide methionine sulfoxide reductase): all samples, FPKM

neo11untrt neo12untrt neo13untrt neo14untrt neo15untrt

0 37.5675.13112.69150.25187.82225.38262.94300.51339.78 0 37.5675.13112.69150.25187.82225.38262.94300.51339.78 0 37.5675.13112.69150.25187.82225.38262.94300.51339.78 0 37.5675.13112.69150.25187.82225.38262.94300.51339.78 0 37.5675.13112.69150.25187.82225.38262.94300.51339.78

expression, by transcript expression, by transcript expression, by transcript expression, by transcript expression, by transcript

10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000

genomic position genomic position genomic position genomic position genomic position

neo16untrt neo17untrt neo18untrt neo3trt neo4trt

expression, by transcript expression, by transcript expression, by transcript expression, by transcript expression, by transcript

10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000

genomic position genomic position genomic position genomic position genomic position

neo5trt neo6trt neo7trt neo8trt neo9trt

expression, by transcript expression, by transcript expression, by transcript expression, by transcript expression, by transcript

10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000 10156000 10160000

genomic position genomic position genomic position genomic position genomic position

FIGURE 2.19: These plots show levels of transcript structures of the differentially expressed isoforms LOC100646534 gene proﬁle between imidacloprid and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 55

There were also differences in expression of LOC100646534 (gene) by production multi different isoform between treated and untreated bees (ﬁgure 2.20; A). This gene correlates with asparaginase activity. asparaginase is an enzyme that primarily catalyze the conversion of L-asparagine (L-Asn) to L-aspartic acid (L-Asp) and ammonia [232] and it is used to treat leukemia blood cancer disease [233]. But little is known about L-asparaginases expressed in mammalian, animal and insect tissues. LOC100644045 methyltransferase-like protein gene and LOC100647513 tRNA methyltransferase 10 homolog A gene showed different expression isoforms at p value 0.003 (ﬁgure 2.20; < B). Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 56

A Control Neonicotinoid

B Control Neonicotinoid

FIGURE 2.20: These plots show levels transcript structures of the of a gene proﬁle between neonicotinod and control samples. Highly expressed transcript depicted in dark red while low expressed transcript depicted in yellow. A; LOC100646534 L-asparaginase (L-asparaginase) and B; LOC100644045 gene (methyltransferase-like protein) Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 57

2.3.5.1 GO analysis

In the list of genes differentially spliced, after FDR correction 7 GO terms were enriched in biological process, 20 GO terms molecular function and 16 GO terms cellular component. For biological process, the ﬁve most represented categories were: protein repair, protein initiator methionine removal, carboxylic acid transport, regulation of rubidium ion transport and ion transport. For cellular component GO terms, the most represented categories were multi-eIF complex and nuclear microtubule. For molecular function GO terms, interestingly, we have found many important functions: toxin transporter activity, peptide-methionine (S)-S-oxide reductase activity, asparaginase activity, peptide antigen binding and gamma-glutamylcyclotransferase activity (ﬁgure 2.21).

REVIGO Gene Ontology treemap

toxin transporter activity

asparaginase activity gamma−glutamylcyclotransferase activity

peptide−methionine (S)−S−oxide reductase activity

peptide antigen binding

FIGURE 2.21: GO term enrichment for molecular function for GO terms (p 0.05) after FDR < correction, associated to genes containing signiﬁcant differentially alternative splicing events. Each rectangle represents a single cluster of closely related GO terms. These rectangles are joined into different coloured ‘superclusters’ of loosely related terms. The area of the rectangles represents the p-value associated with that cluster’s enrichment. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 58

2.3.6 DNA methylation - Expression correlation

We calculated the average percentage of methylation per gene for both differentially expressed genes and non-differentially expressed genes (Figure 2.22), fitting a generalized linear model (GLM) with a quasi binomial error distribution with treatment (control vs imidacloprid) and expression state (DEG vs. non-DEG) as independent variables. For all three groups (CpGs, CHHs and CHGs) there was no significant interactions between the independent variables (interaction model versus main effects only model: CpG: χ2 = -0.0056, d.f. = 1, p = 0.887; CHH: χ2 = -0.00092, d.f. = 1, p = 0.374; CHG: χ2 = -0.0069, d.f. = 1, p = 0.127). For CpGs, non-differentially expressed genes had more methylation than differentially expressed genes (z1,19673=3.875, p<0.001). There was no significant treatment effect on methylation levels (z1,19673=-0.773, p=0.708) (Figure 2.22 A). For both CHHs and CHGs we have found the opposite with no significant expression effect (CHH: z1,19767=2.052, p=0.0876; CHG: z1,19725=1.055, p=0.523), but with imidacloprid treated samples having higher methylation levels than control samples (CHH: z =17.236, p<0.001; CHG: z =9.632, p 0.001) (Figure 2.22 B and C). 1,19767 19725,1 <

I calculated the average percentage of methylation per gene for both differentially expressed isoform (DEI) and non-differentially expressed isoform (non-DEI), ﬁtting a t- test with a proportional of methylation distribution between (DEI vs. non-DEI) as independent variables. In all three groups (CpGs, CHHs and CHGs), there were no sig- niﬁcant differences in methylation proportion between genes DEI and non-DEI . CpG; t = -0.60572, df = 26.321, p-value = 0.5499. CHG; t = -1.3264, df = 26.61, p-value = 0.196. CHH; t = -0.85511, df = 25.415, p-value = 0.4005.

To have a more ﬁne scale understanding of the correlation between methylation and expression, we plotted mean proportion of methylation per gene against ranked expression level (log10 fpkm per gene) in 100 bins (from low to high) (Figure 2.23) ﬁtting a linear model with treatment and expression level as independent variables. For all Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 59

FIGURE 2.22: Average percentage of methylated CpG (A), CHG (B) or CHH (C) per gene. Control samples are in grey and Neo treated samples in white. Differentially expressed genes (DEG) and non differentially expressed genes (nonDEG) are plotted separately. Dots represent genes. three groups (CpGs, CHHs, CHGs) there was no signiﬁcant interaction between expression’s and treatment’s effects on methylation (interaction model versus main effects only model: CpG: F1,189 = 1.0347, p = 0.3104, CHH: F1,189 = 0.0304, p = 0.8617,

CHG: F1,189 = 0.621, p = 0.4317). We found a signiﬁcant effect of expression on methylation in all contexts except for CHG (CpG: F = 281.654, p = 2 x 10-16, CHH: F 1,189 < 1,189 -4 = 11.939, p = 6.787 x 10 , CHG: F1,189 = 0.0446, p = 0.8329). Imidacloprid treated bees had comparable levels of CpG methylation to control bees (F1,189 = 1.8125, p = 0.1798). However imidacloprid treated bees had higher levels of CHH and CHG methylation than controls (CHH: F = 89.189, p 2 x 10-16, CHG: F = 66.3945, p = 4.938 x 1,189 < 1,189 10-14). Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 60

0.0025

0.0020

0.0015

0.0010

0.0005 Control Neo Proportion CpGs per Gene of Methylated

0.0000 0 25 50 75 100 Expression Rank (Low to High)

0.00046

0.00044

0.00042

0.00040

0.00038

Control 0.00036 Neo Proportion CHGs per Gene of Methylated

0.00034 0 25 50 75 100 Expression Rank (Low to High)

0.0018

0.0017

0.0016

Control Neo Proportion CHHs per Gene of Methylated

0.0015 0 25 50 75 100 Expression Rank (Low to High)

FIGURE 2.23: The proportion of methylated CpGs (A), CHHs (B) and CHGs (C) is plotted against gene expression rank. One hundred “bins” of progressively increasing level of expression were generated and genes with similar level of expression have been grouped in the same bin. Solid lines represent control samples and dotted lines imidacloprid treated samples. The grey shading represents 95% conﬁdence intervals. Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 61

2.4 Discussion

I found numerous genes which show differential expression or methylation between bees treated with ﬁeld realistic doses of the neonicotinoid imidacloprid and control bees. I found CpG methylation much more focussed in exons compared to CHH or CHG methylation. High CpG methylation was associated with highly expressed genes. Neonicotinoid treated bees had higher levels of non-CpG methylation.

The preponderance of differentially expressed genes associated with synaptic transmission are to be expected, given that I used brain tissue and given the known target effects of neonicotinoids. The effects on metabolic pathways was also recently found in honeybees [234]. It could be argued that due to the intensive energy demands of the brain, negative effects on metabolic pathways could affect brain function and therefore behaviour [235]. Increased activation in the glutathione metabolic process as result of neonicotinoid in this study corroborates these earlier ﬁndings [236].

On the question of the effects of neonicotinoids on alternative splicing, this study found that imidacloprid also altered transcript levels of genes related to oxidase reduction pathway and genes responsible on transport toxin and sugar metabolism in the fat body into the hemolymph. Interestingly, Trehalose gene showed differential methylation in the exon level ( ﬁgure 2.4, Tret) and expressed isoforms (ﬁgure 2.18 A). Learning and memory in honey bees is associated sugar metabolism [237]. Bees prefer to eat more sucrose when foods contain neonicotinoids [238]. It can therefore be hypothesised that the worker bumble bee has a similar response to neonicotionod and this causes the increased treholase levels in the blood.

I have found a number of differentially methylated Cs (79 CpGs, 86 CHHs, 16 CHGs). Some of the genes containing these differentially methylated Cs are involved with neuron signalling or brain and eyes development (append). Moreover, only a few genes (see table 2.2) showed differentially methylated Cs in different genomic contexts (CpGs, CHHs or CHGs). This suggests different functions and/or targets for CpG and Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 62 non-CpG methylation. Gene ontology analysis of the biological functions of genes containing differentially methylated Cs returned a complex picture with some indication of function related to neuron signalling, neuron differentiation and acetylCoA activity.

Although I found a number of differentially methylated genes, imidacloprid seen to have no effect on the overall levels of CpG methylation (see ﬁgure 2.22 and 2.23). As expression rank increases the level of methylation increases till it reaches a peak and levels off at higher expression ranks (see ﬁgure 2.23A). This pattern is similar to those found in other social insects for CpG methylation and expression [239].

Gene body methylation has been shown to play role in histone modifications and nucleosome stability in mammals and also differential DNA methylation levels have been specifically shown to be higher for alternative exons [240]. In insects, DNA methylation influencing nucleosome stability and positioning, DNA methylation may contribute to the definition of exons during the transcription and splicing of mRNA [241]. I conclude that change in expression and methylation in this study supports the concept that epigenetics holds substantial potential for furthering our understanding of the molecular mechanisms of neonicotinoids health effects which can exert profound influence on the development and function of the future individual.

Non-CpG methylation plays a role in gene silencing in flowering plants [242] and to a lesser extent, in mammals [243]. To the best of my knowledge its role has never been directly studied in insects. Wang et al. [244] stated that the jewel wasp’s (Nasonia) genome lacked non-CpG DNA methylation. We found low but measurable amounts of non CpG methylation in the buff-tailed bumblebee. The discrepancy is possibly due to our samples consisting of whole brains compared to Wang et al. 244 whole body samples. Non-CpG methylation is found at higher levels in mammalian brains compared to other tissues [39]. We found that CHH and CHG methylation was uniformly spread throughout genes, (see figure 2.2). This is in stark contrast to CpG methylation’s preponderance in exons, a result found in multiple invertebrate species [157]. This difference in distribution between CpG and nonCpG methylation again suggests Chapter 2. The epigenetic effects of neonicotinoids on bumble bees 63 a different molecular function for them in bumblebee biology. Imidacloprid treatment consistently raised the level of non-CpG methylation (see figure 2.23 A). In mammals, it has been hypothesised that neurons could use non-CpG methylation to control their gene expression during critical periods [245].

Recently, it has become clear that epigenetics can play a role in the interplay between man-made chemicals and natural ecosystems, and their constituent species [153]. Hy- menopteran insects (ants, bees and wasps) are important emerging models for epigenetics [163, 246–248]. Given that we have established epigenetic effects due to a neonicotinoid, one interesting future area of enquiry would be to ask if there are any intergenerational effects due to neonicotinoids. Intergenerational epigenetics is deﬁned as effects found in the offspring generation (F1) due to direct exposure to the stressor of the parental (F0) and/or the developing germ cells. There are examples of intergenerational effect due to pesticide exposure [152]. For example, exposing rats to the fungicide vinclozolin leads to effects on male reproductive functions into the F4 generation [249]. A similar ﬁnding for neonicotinoids on bees would be important in informing future legislation controlling these important pesticides.

2.5 Collaborative work statement

Assistant prof Dr. Eamonn Mallon and myself designed the project. Bee husbandry, brain dissection, DNA and RNA extraction , methylation analysis, differential expression analysis, alternative splicing analysis, enrichment analysis and KEGG pathway analysis were carried out by myself. Correlation analysis between methylation and expression were completed together by Dr. Mirko Pegoraro, Hollie Marshall and myself. List of GO terms was prepared by Alune Jones and he assisted me in the enrichment analysis. Chapter 3

The neonicotinoid, Imidacloprid affects gut bacterial community in Bombus terrestris workers.

64 Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 65

3.1 Introduction

The gut microbiome in both humans and animals is recognised as playing an important role in health and quality of life [250]. It can beneﬁt hosts in numerous ways including by helping to digest food, detoxifying harmful molecules, providing essential nutrients, protecting against invasion by pathogens and parasites, and by modulating development and immunity [1, 251–253]. On the other hand, the composition of the gut microbiome is susceptible and can be inﬂuenced or disturbed during the course of physiological and behavioural changes. Host development and morphogenesis [254], geographic location, season, age and diet[255], stress from pathogenic and insecticides [256, 257] have all been shown to affect gut microbiomes.

So far in my thesis I have shown that neonicotinoid insecticides effects on many genes related to brain and immunity system (see chapter 2). Could neonicotinoids have an effect o the gut microbiome as well? There are several lines of evidence suggesting this. Generally, there is a link between gut microbiome and function of the central nervous system [132]. Other pesticides, for example, the fungicide Pristine adversely affects the relative abundance of Lactobacillus sp. Firm 4 and Firm 5 in the guts of honey bee colonies [133]. In a previous study, exposure to neonicotinoid pesticides, due to different environmental landscape oilseed rape coverage, was associated with changed microbiome composition in the honey bees [137]. One possible mechanism to explain this could be the known impaired foraging efﬁciency of bumble bees treated with neonicotinoids [115] affecting the gut micobiome. Another is the known link between social behavior, also affected by neonicotinoids, and the bacterial community variation in bees [258]. Studying the relationship between gut bacteria and neoincotinoids will help us understand the deeper connections between neonicotinoids exposure and the health of bees.

The digestive tract in both bumble bees and honey bees is known to harbour a wide variety of commensal or beneficial bacteria [1]. The bee’s digestive tract is divided into Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 66 four major compartments: crop (honey stomach), midgut, ileum and rectum and approximately 10 billion bacterial cells exist in the bee’s gut (figure 3.1)[1]. Each compart- ment is dominated by a distinctive set of specific bacterial species and largely restricted to the hindgut [1, 259]. Based on 16S rRNA gene profiling, bees’ guts are dominated by eight or nine stably associated gut bacterial taxa (phylotypes) [260].

FIGURE 3.1: The digestive tract structure of honey bee and bumble bee species with bacterial localization in the different compartments: the crop, midgut, pylorus, ileum and rectum ( ﬁgure taken from Kwong and Moran [1]) Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 67

Six phylotypes belong to gram negative bacteria and these are all largely restricted to bee hind-guts. Five are within the phylum Proteobacteria; these include one betaproteobacteria within the genus Candidatus name Snodgrassella alvi (family Neis- seriaceae) [1] , two closely gammaprotobacteria within the genus Candidatus name gammaproteobacterium Gilliamella apicola (family Orbaceae) [261], and two other species clusters are from distantly related clusters within genus alphaproteobacteria, initially called Alpha1 and Alpha2; newly described as species Frischella perrara in honey bees [262]. The sixth phylotype, Apibacter adventoris is a member of the phylum Bacteroidetes have been isolated at low abundance in guts some species of honey bee and bumble bee [263, 264].

Three gram-positive genera are isolated in bee guts; include two clusters within genus lactobacillus (Lactobacillus Firm-4, Lactobacillus Firm-5 (family Lactobacil- laceae, phylum Firmicutes)) [265] and one Bifidobacterium sp. (family Bifidobacteri- aceae, phylum Actinobacteria) [266]. Bifidobacterium species (Bifidobacterium actinocoloniiforme, Bifidobacterium bohemicum and Bifidobacterium bombi) are predominantly isolated from the rectum of bumble bees’ guts [267, 268].

There are several strains of Lactobacillus; one group belong to species Lactobacillus kunkeei , two groups belong to Lactobacillus buchneri and Lactobacillus delbrueckii. Seven novel species have recently been detected and cultivated from the gut of the honey bee; these include Lactobacillus mellis, Lactobacillus apinorum, Lactobacillus mellifer, Lactobacillus melliventris, Lactobacillus kimbladii, Lactobacillus helsingborgensis and Lactobacillus kullabergensis. Praet et al. in 2015, isolated two other novel lactic acid bacterial species from the bumble bee gut named Lactobacillus bombicola and Weissella bombi.

Lactic acid bacteria and their role in producing lactic acid and antibacterial peptides seem particularly important in bee health [264, 270, 271]. Lactobacillus strains exert probiotic properties in insects [256] and Lactobacillus is able to reduce organophosphate and neonicotinoid pesticide absorption and toxicity to Drosophila melanogaster Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 68

[253, 256]. Lactobacillus kunkeei presense is associated with reducesd infection probability of American foulbrood disease and intensity of the common honeybee pathogens Paenibacillus larvae and Nosema ceranae [272]. Apibacter, Lactobacillus Firm-5, and Gilliamella protect Bombus impatiens against Crithidia infection [273]. Biﬁdobac- terium reduces the infection probability and intensity of the common trypanosome gut parasite Crithidia bombi. Together, these studies based on 16S rRNA approach indicate that bees gut bacterial community aids the bee’s immune system.

16S ribosomal RNA (rRNA) is part of the 30S small subunit of a prokaryotic ribosome and it is highly conserved between different species of bacteria and archaea [274]. In bacteria, the 16S rRNA sequence contain hypervariable regions that are widely used for phylogenetic studies, identification, classification and quantification of bacterial communities in complex environmental samples [275]. The length of the 16S rRNA subunit is 1500, containing nine hypervariable regions (V1-V9) ranging from about 30- 100 bp long (figure 3.2)[276]. V4-V6 are optimal sub-regions for the design of universal primers for representing the full-length 16S rRNA sequences in the phylogenetic analysis of most bacterial phyla [277]. The V4 region has been previously used in bees gut micrbiome projects [137, 273, 278].

Several sophisticated bioinformatic tools and pipelines have been developed to classify and quantify 16S rRNA sequences of bacterial communities; including QIIME (Quan- titative Insights Into Microbial Ecology), mothur, MG-RAST (Metagenomics - Rapid Annotation using Subsystems Technology), Genboree, EzTaxon, Pheonix2, METAGE- Nassist, MEGAN, VAMPS, SnoWMan, CloVR-16S, the RDPipeline (Ribosomal Database Project Pipeline), Vegan, ade4, and ape [279]. Qiime is the quickest and produced the most accurate quality control, clustering of similar sequences, assigning taxonomy, calculating diversity measures and visualising results [280]. As regards taxonomic classification and assigning the most likely taxonomic lineages, all these tools compare and align raw reads against a defined reference database, such as Greengenes [281], NCBI [282, 283], RDP [284], or SILVA [285], the SILVA database generally yielded a higher re- call than other databases [286]. FIGURE 3.2: Illustration of nine hypervariable region in 16s rRNA gene were identified among bacteria. Red sections are start sites of primers (consiver ; green and orange raws are name of primers, light green (blue) sections are length of PCR amplification (DNA sequence). The figure refers to 16S rRNA and 16S rRNA Gene in the EzBioCloud Help center (http://help.ezbiocloud.net/ 16s-rrna-and-16s-rrna-gene/) . Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 70

The aim of this study was to investigate the effect of field reaslistic doses of the neonicotinoid imidacloprid on the stable gut bacterial community of non reproductive workers in the bumble bee Bombus terrestris. I produced and analysed thirty libraries of 16S rRNA metagenomic data from biological samples from the mid and hind guts of fifteen neonicotinoid exposed non-reproductive workers and fifteen control non- reproductive workers. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 71

3.2 Methods

3.2.1 Neonicotinoid toxicity test

To assess the in vitro bacterial toxicity of our selected neonioctinoid, various concentrations of the neonicotinoid imidacloprid were prepared (0 ppb, 5 ppb, 7.5 ppb, 10 ppb and 20 ppb) in sugar water.

Lysogeny broth (LB), Lysogeny agar (LA) media, and isolates of Staphylococcus aureus and Escherichia coli were obtained from lab 121 - Genetics and Genome Biology department - Leicester university. Both species were cultured on agar plates and after 24 h, single colonies was selected and then inculcated in LB and incubated at 25 °C for 24 h. After that, 30 LA plates were prepared. 10 µl of bacterial suspension were added to each LA plate (x15 Staphylococcus aureus - x15 Escherichia coli) then immediately suspension were spread out over the surface of agar using a sterile cotton swab. Plates were left for 10 minutes, then 10 µl of sugar water + imidacloprid (concentration x3) were added in the centre of plates, after that, plates were left for 10 minute near ﬂame to dry out. After an overnight incubation, the bacterial growth around place of sugar solution were measured.

3.2.2 Bee husbandry, imdicloprid exposure and gut sampling

Five colonies of Bombus terrestris audax were reared as previously described in section 2.2.1. In each colony, 30 callow workers (less than 12h old - day 0) were collected, 15 bees were treated with nectar and pollen containing 10 pbb of imidacloprid and 15 workers were fed untreated nectar and pollen as a control. In total I had 3 replicates per treatment group per colony, 5 bees in each replicate set.

Each callow worker was individually tagged on the top side of the thorax using the Queen Marking Kit (E.H. Thorne Beehives Ltd) and Uni Posca PC - 5M water based Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 72 markers. Different colours were used to label and identify the 5 bees in each box: Red (R), Green (G), Blue (B), Yellow (Y) and White (W). As soon as the labelling was completed the bees were returned immediately back to boxes containing 3 adult workers. This was to share the microbiome between individuals. Callow workers have less core microbiome (lee than 100 bacterial cells) compared with adult workers (109 bacterial cells) [1]. Faeces was collected from all treatment groups on day 6 and 7. After that in day 8, guts were dissected out from the bees using aseptic techniques. Bees were anaesthetized in ice for 15 minutes. First, the surface of their bodies were sterilized by submerging bees in 75% (v/v) ethanol for 2 minutes. To remove ethanol, bees were submerged into distilled water for 2 minutes and then placed to dry on sterilised tissue paper. The hind and mid gut was pulled outwards gently and then kept in - 20 ◦C for later use. The faeces samples have been similarily stored, but were not used in this study.

3.2.3 DNA extraction and library preparation

Two guts of non-reproductive workers were pooled according to replicate set, colony and treatment type. DNA was isolated from a total of 60 (30 control - 30 neonicotinoid) bees using the Fast DNA ® SPIN Kit for Soil and the FastPrep Instrument (MP Biomed- icals, Santa Ana, CA), as per manufacturer’s instructions. DNA samples were stored at 20 °C for later DNA sequencing. The concentration of genomic DNA was measured using a Qubit® dsDNA BR Assay Kit (ThermoFisher Scientiﬁc, USA).

Thirty libraries of paired-end reads were generated with Illumina HiSeq/MiSeq platform at BGI Tech Solution Co., Ltd. (Hong Kong). Ampliﬁcation hypervariable region (V4) 16S rRNA sequencing protocol was performed with reads length 250 * 2 (forward and reverse reads) bp using forward Primer : 515F 5’ GTGCCAGCMGCCGCGGTAA 5’ and reverse primer 806R 5’ GGACTACHVGGGTWTCTAAT 3’. Reads with sequencing adapters, N base, poly base etc. were ﬁltered out by BGI. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 73

3.2.4 Bioinformatic analysis

All analysis were performed using the QIIME 2 Core v2018.2 pipeline [287]. First, sequence data (FASTQ ﬁles, includes Read1 and Read2) were imported into a QIIME 2 artifact. This was done by using demultiplexing sequences function with default command; qiime tools import and –input-path pe-64-manifest and –source-format PairedEndFastqManifestPhred64. Then, a sequence quality control was conducted for all samples using a DADA2 v1.8 pipeline [288] implementing the default function; dada2 denoise-paired plugin in QIIME 2. The quality proﬁles of forward and reverse reads were visualized in qiime2 view browser. The forward reads were of generally good quality with quality decreasing in the last ten nucleotides. These ten nucleotides were removed with function –p-trunc-len 240. The reverse reads were of worse quality at the last 80 nucleotides. I trimmed the last 88 nucleotides with the function; –p-trunc-len 162.

The forward and reverse reads were merged by overlap of at least 100 bases by identity in the overlap region. A feature table of sequence variants (Operational Taxonomic Units (OTUs)) was generated with the standard cutoff of 97% similarity between sequences using; qiime feature-table summarize plugin in Qiime2. Sequence table (OTU) is a matrix with rows corresponding to samples, and columns corresponding to the sequence variants (number OTUs or species or phylotype). Features that were present in only a single sample and with less than 5 sequences were ﬁltered from the feature table.

Phylogenetic diversity analyses metrics were prepared to perform Faith’s Phylogenetic diversity and weighted and unweighted UniFrac [289], which depends on a rooted phylogenetic tree relating the features (OTU) to one another. A multiple sequence alignment of the sequences (feature sequence) was performed to create a aligned sequence using the mafft program [290] with a default function qiime alignment mafft plugin in qiime2. Some positions may have passed through the denoised quality control and Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 74 added noise to a resulting phylogenetic tree. So, I masked the alignment reads to remove positions that were highly variable using; qiime phylogeny fasttree plugin in qi- ime2. Features assigned as mitochondria and chloroplast were removed using qiime taxa ﬁlter-table pluging. Finally, qiime phylogeny midpoint-root function was applied to place the root of the tree at the midpoint of the longest tip-to-tip distance in the unrooted tree.

Alpha and beta diversity analysis were implemented in QIIME2 through the q2- diversity plugin.

Different types of Alpha diversity was performed to measure the level of diversity within individuals in both control and imidacloprid samples. The analysis including Shan- non’s diversity index (a quantitative measure of taxon (bacterial species) richness depending on OTU table), Evenness or Pielou’s Evenness (it is same as Shannon’s index, accounts for both abundance and evenness of the taxon present) and Faith’s Phylo- genetic Diversity (a qualitative measure of taxon richness and it is expressed as the number of tree units (clades) which are found in a sample).

Beta diversity measures the level of diversity or dissimilarity between condition (control vs imidacloprid). I calculated several measures of beta diversity including Jaccard distance (a qualitative measure of community dissimilarity in phylogenetic diversity (evolutionary diversity in DNA sequences) based on presence/absence by determin- ing what percent of bacterial identiﬁed were present in both groups), Bray-Curtis distance (a quantitative measure of community dissimilarity in number of species based on abundance species (OTU)), unweighted UniFrac distance (a qualitative measure of community dissimilarity based on phylogenetic diversity) and weighted UniFrac distance (a quantitative measure of taxon dissimilarity that directly accounts for differences in relative abundances).

Sampling depths across different samples were adjusted to 1120 sequences using –p- sampling-depth, This was performed because diversity metrics are sensitive to different sampling depths across different samples. Principal coordinates analysis (PCoA) Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 75 was performed on each beta diversity metric to highlight the separation of samples based on distance between groups. Permutational multivariate analysis of variance (PerMANOVA) on beta diversity matrices were used to test differences between imidacloprid and control.

3.2.5 Taxonomy analysis

Classiﬁcation and taxonomic analyses were conducted to explore the the organisms that were present in the samples. The sequences in FeatureData[Sequence] were aligned against bacterial references of the Silva 99% database trimmed to the V4 region using feature-classiﬁer classify-sklearn in QIIME2. Taxonomic compositions of samples were visualized by producing bar plots using qiime taxa barplot function.

Phylogenetic trees were constructed from assigned sequences for all sequence variants in control and imidacloprid samples separately. This was to visualize evolutionary relationships among various species. First, in order to choose the best model for alignment between sequences (sequence variants), MEGA7 detected General Time Reversible (GTR) as best alginments methods for my sequences. ClustalW alignment (with multiple alignment) was implemented for the examined sequences by Geneious9 software. Then, phylogenetic trees were reconstructed by General Time Reversible (GTR) as a best substitution model using Bayesian inference of phylogeny (MrBayes 3.2.6).

3.2.6 Differential abundance analysis

Features (species) were removed if they had a total abundance less than 10 % across all samples. This was carried out using feature-table filter-features plugin in qiime2. Differential abundance analysis was conducted using Gneiss pipeline in qiime2. The steps were implemented as following; (1) I defined partitions (proportion) of microbes (species) in each sample to construct balances, (2) grouped species together (hierarchical clustering) based on how often they co-occur with each other (proportional count Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 76 of sequence), all these two steps were implemented using a qiime gneiss correlation- clustering plugin, (3) I performed the isometric log ratio transformation to compute the log ratio between groups each node in the tree. (4) then, a linear regression (multivariate response linear regression) was performed on each balance separately using qiime gneiss ols-regression plugin. (5) all the coefficient p-values for each of the balances was visualized in heatmap, using qiime gneiss dendrogram-heatmap plugin. Fi- nally, the log ratio of top two balances were plotted in a boxplot using qiime gneiss balance-taxonomy plugin, also this show that how many species in denominator and numerator, and then visualize the differences between the control and imidacloprid groups. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 77

3.3 Results

3.3.1 Neonicotinoid anti bacterial toxicity

To test neonicotinoid in vitro toxicity against bacteria, I examined the effect of several doses of neonicotinoid on Escherichia coli and Staphylococcus aureus using a standard protocol. The results showed that there were no zone of inhibition of these bacteria around imidacloprid at low dose (5 ppb), ﬁeld realistic (10 ppb) and lethal dose (20 ppb).

3.3.2 Bioinformatic metagenomic analysis

A total of 3,092,487 sequences (minimum 16,907, mean 103,082.9, median 91,818.5 and maximum 337,267) with length 2 x 250 bp were obtained from 30 libraries (5 colonies * 3 replicates * 2 condition (control and imidacloprid)). Sequences have been generated on an Illumina MiSeq for the hypervariable region (V4) of 16S rRNA subunit. Forward reads for all samples showed higher quality with the large majority of bases having quality scores above Q30 (Inferred Base Call Accuracy, 99.9%). Reverse reads showed good quality scores (Q30) while worse at the last 85 sequence below Q15.

The bad quality sequences were removed leaving 2,233,574 sequences (see Table 3.1) to construct the feature table. Feature table (OTU picking) from all 30 samples at the 97% similarity between reads gave 129 (83 control and 97 imidaloprid) amplicon sequence variants (features or OTU or phylotypes or species) across all samples. Sequences of 32 features mapped to chloroplast and mitochondria DNA sequences and 37 features presented in one sample and have less than 5 sequences. All theses features were excluded in further analysis. Overall, 60 features were used for further analysis (alpha, beta and composition diversity). Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 78

Condition No. samples Min frequency Max frequency Mean, Total control 15 13,060.0 201,397.0 61,322.3 919,835 Neonicotinoid 15 37,534.0 226,663.0 84,451.9 1,266,779

TABLE 3.1: Shows summary of sequences were retained after trimming and removing low quality reads

3.3.3 Diversity Indices

I checked if imidacloprid altered the number and proportion of bacterial species in individuals. I investigated differences in alpha diversity across all samples (imidacloprid and control samples) using three alpha diversity measurements. The results showed no differences in the alpha diversity in any of these three analyses (Faith’s phylogenetic diversity; Kruskal-Wallis (pairwise) test; q value = 0.480682 (figure 3.3), and measure of community evenness for all samples; Kruskal-Wallis (pairwise) ; q value 0.575511, (figure 3.4). Shannon evenness analysis (figure 3.5) showed that, there were strong similarities in taxon (OTU) between samples in both group (q value = 0.205842). I concluded that no association between imidacloprid exposure and phylotypes richness and evenness, suggesting that imidacloprid was not a strong driver of variation in bumble bees microbial communities.

Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 79 Faith’s Phylogenetic Diversity Faith’s

Control Imidacloprid

Condition

FIGURE 3.3: Boxplots showing Faith’s phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of imidacloprid and control samples. X axis indicates samples; Y axis indicates proportional of similarity bacterial species between samples. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 80

FIGURE 3.4: Boxplots show measure of community evenness (alpha-diversity indexes) in total difference proportions of bacterial species present in bumble bees. X axis indicates samples; Y axis indicates proportional of similarity bacterial species between samples . In a site with low evenness indicates that a few species dominate in the sam-

ple

Phylogenetic Diversity Shannon

Control Imidacloprid

Condition

FIGURE 3.5: Boxplots showing Shannon phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of imidacloprid and control samples. X axis indicates control and imidacloprid groups; Y axis indicates Shan- non indexes of similarity bacterial species between groups. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 81

Beta diversity analysis were performed to measure the overall change in the whole community (imidacloprid vs control). This was performed by comparing multivariate statistics (weighted UniFrac distance, unweighted UniFrac distance, Bray Curtis distance and Jaccard distance) . In all analysis, I did not find a strong difference (distance) between imidacloprid and control groups. Weighted UniFrac distance matrix was performed to test if there were differences in abundance of dominant species between groups. As shown in PCoA Plot (figure 3.6), there were no difference in gut bacteria community between imidacloprid and control group. Unweighted UniFrac distance was performed to emphasis differences in phylogenetic lineages between groups. Result showed no difference in rare species between groups (Pairwise permanova (pseudo-F) method, q value = 0.861, figure 3.7).

FIGURE 3.6: 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points control samples; red points neonicotinod samples. Distances between samples were calculated using weighted UniFrac distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 82

FIGURE 3.7: 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points control samples; red points neonicotinod samples). Distances between samples were calculated using unweighted UniFrac distance matrix. Samples stay close to each other means that those samples have communities with very similar overall phylogenetic trees .

I graphically represented other similarities between individuals in groups and differences between groups using PCoA based on Jaccard similarity index. This was generated by depending on rare species (presence or absence) in the data in both control and imidacloprid groups. Results showed some separation between some samples in control group (blue points) and imidacloprid samples (red point) in phylogenetic distance dissimilarities (ﬁgure 3.8). Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 83

FIGURE 3.8: 3D PCoA Plot showing a sample-by-sample distance based on Jaccard distances in bacterial communities, showing separation of samples by sample type between control and neonicotinoid. Each point represents one of the samples (green points control samples; red points imidacloprid samples. Distances between samples were calculated based on presence/absence. Samples close to each other means that those samples have percentage species with very similar overall phylogenetic trees. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 84

3.3.4 Differential abundance analysis

Hierarchical correlation clustering was generated according to partitions of microbes that commonly co-occur (proportion) with each other across all samples. The data was split into 10 partitions (Figure 3.9, Y0 - Y9). The first partition (root of tree) divided into two clades. The first clade included one species Lactobacillus (figure 3.10; A) and second clade included five species (Lactobacillus, Snodgrassella, Methylobac- terium, Streptococcus and Lepttrichia) (figure 3.10; B). As shown in the heatmap tree (Figure 3.9), there were no significant differences between groups in the abundance between groups in 59 species that them abundance were correlated to each other (figure 3.10; A) while the abundance the first top species (figure 3.10; A) was extremely difference between groups.

According to the NCBI and Silva database, the sequence of this species is Lactobacillus helsingborgensis (identity 100%., query cover 100%, GenBank: MG958540.1). Previ- ously this species was isolated from guts of the honey bee Apis mellifera [265].

The ratio of Lactobacillus helsingborgensis (figure 3.10; B) vs all species (figure 3.10 A) was roughly 50 times greater in imidacloprid condition compared to control condition (figure 3.11). So Lactobacillus may be extremely affected by imidacloprid treatment in our experiment. It is possible that Lactobacillus helsingborgensis increased significantly between these treatments. So, the species in the y0numerator (figure 3.10; B) on average are less abundant in the imidacloprid group compared with the control group and the species in the y0 denominator (figure 3.10; A) on average was greater in imidacloprid groups compared to control group (figure 3.11). Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 85

control neonics y0 y1 y2 y3 y4 y5 y6 y7 y8 y9 condition

FIGURE 3.9: Heatmap of the log of the coefﬁcient p-values for each of the balances. The rows of the heatmap represent samples and the columns of the heatmap represent balances. The colour scale represents microbial abundances at p value linear regression on the balance by computing the log ratios between species. Each of the tips corresponds to a taxon, Y represent split the data into partitions, for example Y0 is partitions between long branches (sub trees) and short branches (sub trees). light red in Yo balances is numerators for each balance and dark red is denominators. . Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 86

y0numerator taxa (59 taxa)

D_5__Lactobacillus

D_5__Snodgrassella

D_5__Methylobacterium

D_5__Streptococcus

D_5__Leptotrichia

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

y0denominator taxa (1 taxa)

D_5__Lactobacillus

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Number of unique taxa

FIGURE 3.10: Shows number of unique taxa (phylotype) in both numerator and denominator partitions

y0 y0 = ln numerator y0denominator

control log ratio

treated

8 6 4 2 0

FIGURE 3.11: Boxplots showing balance taxa summary, x axis represents average the log ratio for each sample, at each balance, calculated the isometric log ratio transform ; y axis control and treated neonicotionoid samples. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 87

3.3.5 Taxonomy and relative abundance

The four most abundant phyla in our 16S rRNA gene sequences (excluding unclassified bacteria, cynobacteria (Chloroplast)) were: Betaproteobacteria, Gammaproteobacte- ria, Actinobacteria and Firmicutes (figure 3.12). These phyla cover the core gut bacteria as previously found in other adult bumble bee species [261]. The first three phyla’s relative abundance in bees exposed to imidacloprid were slightly less compared with untreated bees. The Firmicutes phyla showed a modest decrease in the control group.

Relative abundance of Betaproteobacteria - order Neisseriales in all samples was 37.45 % (control 38.1 %; imidacloprid 36.8 %); Gammaproteobacteria - order Orbales was 30 % (control 31.4 %; imidacloprid 29.6 %); Actinobacteria - Biﬁdobacteriales was 10.6 (control 9.8 %; imidacloprid 11.8 %) and Firmicutes - order: Bacillus 18.95 (control 17.1 %; imidacloprid 20.8 %). On overage, 0.73 % of bacterial community assigned to phylum Cyanobacteria (Chloroplast) and less than 0.05 bacterial community belong to phylum; Bacteroides.

The most representative genera were : (1) beta proteobacterium (Snodgrassella) (sequence counts: 359,817 control - 467,570 imidacloprid); (2) Bifidobacterium was detected in control samples with 109742 sequences and 132021 sequences in neonicotinoid. This result showed the abundance of Bifidobacterium increased in bees exposed to imidacloprid; (4) Six Gammaprotobacteria species were identified in control samples and seven species in imidacloprid samples (285,568 control - 381,356 ); (5) 25 different species of alpha protobacteria were identified in both control and imidacloprid samples; (4) other unique features were identified as Lactobacillus bombi and Lacto- bacillus bombicola with frequency 270,244 and 107,518 sequences respectively in all 30 samples. 247,038 sequences of DNA were Bifidobacterium sp. and 890 sequences of Ralstonia sp bacteria were detected in all samples; (6) 285,568 sequences in control samples assigned belong to phylum Actinobacteria with eight different species and 6 species with sequences frequency 132021 in imidacloprid samples. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 88

FIGURE 3.12: Taxonomic level plot based on the 16S amplicon sequencing, the top phylum and sub phylum, calculated according to total relative abundance across the sample set. x axis is sample - y axis is relative abundance. The minor phylum which them abundance less 1 % are no showing in plot. grey bars phylum Actinobacteria, green phylum Firmicutes, orange sub phylum Gammaprotobacteria, blue sub phylum Betaprotobacteria. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 89

3.3.6 Phylogenetic tree

A Neighbor-joining cladogram tree was constructed in Geneious 9 from 16S rRNA bacterial sequences (67 control OTUs - 72 OTUs in samples treated with imidacloprid). Analyses of the overall community profile indicated that gut bacteria in our bees was divided into six clades with 94% bootstrap value: (figure 3.13). Clades of Firmicutes and Actinobacteria are grouped together and close to clade alphaprotobacteria (figure 3.13 dark blue colour). Clade of Beta and gamma protobacteria in control group close to alpha protobacteria clade which are light blue colour (figure 3.1 B). In imidacloprid groups, the Clade of Beta and gamma protobacteria are close to Bacteroides clade and both these are close to light blue colour of alpha protobacteria (figure 3.1 B).

As shown (figure 3.13) alphaprotobacteria is divided into two groups, according to the NCBI database alpha species in the light blue group assigned as uncultured bacterium, in Silva database assigned as Family: Anaplasmataceae; genus Wolbachia. Se- quences in other branches which coloured dark blue were classified under the phylum Alphaprotobacteria as well, one species under the order; Sphingomonadales, two species in family: Caulobacteraceae, and others with genus; Acidocella, Rhodobacter- aceae and genus Bombella intestini, two species of Bradyrhizobium, Cladosporium cla- dosporioides. In total, there were 13 species of alphaprotobacterium in imidacloprid samples and 12 species in control samples. Campylobacter species belong to class Ep- silonproteobacteria and Acidobacter species were observed in imidacloprid samples. As shown in (figure 3.13) these two species cluster with the phylum Alphaphylobacte- ria. In bees, all species assigned as alphaprotobacteria are called Alpha1 and Alpha2. Several other species are closely related to alphaprotobacteria previously have been found in guts of other insects [1]. For instance, the Alpha2 species, which appears to include acetic acid bacteria, such as Acetobacter and Gluconacetobacter species, these phylotypes are found in numerous insects, including bees, mosquitoes, flies, leafhop- pers and mealybugs [259]. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 90

The phylotypes corresponding to Gamma and beta protobacteria in our samples ( control = 7 species - neonicotinoid = 14 species) were divided into two groups (beta and gamma). One species out group belong to phylum Chlamydiae in imidacloprid and in control group, sequences of one species belong to phylum Verrucomicrobia was out group and matches equally well with Gamma species (in sister with one species of Gamma).

The sequences assigned to phylum Firmicutes was classified into two branches, Lac- tobacillales and Bacillales. Lactobacillales include many species in the family: strep- tococcaceae, genus :Lactobacillus. Bacillales species were Planococcaceae, Gemel- lales, Veillonella, Paenibacillaceae. one species of Finegoldia was detected in control samples and one species of Paenibacillus larvae in two samples (neo 13 and 14) was detected in imidacloprid samples. Sequences assigned as Paenibacillus larvae were found in both NCBI and Silva database. This is first time this species has being detected in bumble bees and it is known as a destructive and a notifiable pest of honey bee colonies in the UK. According to Animal and Plant Health Agency, American foulbrood disease can lead to the death of infected colonies [291]. A B

FIGURE 3.13: Neighbor joining phylogenetic tree showing the diversity of bacteria in control groups (A) imidacloprid groups (B). Alphaprotoacteria = blue, Actinobacteria = brown, Gamma and beta protobacteria = red; Firmicutes = green. Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 92

3.4 Discussion

Exposure to neonicotinoid insecticides presents diverse risks to bumble bees [115, 128, 292], but its effects on gut microbiota are unknown. To our knowledge, this is the ﬁrst study that has characterized the relationship between insecticide and gut microbial community of bumble bees using 16s rRNA sequencing. Our analysis of the gut microbiota of a non reproductive workers exposed to ﬁeld dose (10 ppb) of imidacloprid showed little or no strong impact on the bacterial composition.

The PERMANOVA analysis revealed no significant effect of the imidacloprid treatment on the distribution of bacteria among samples. The diversity of bacteria within bumble bee individuals, as measured by the Shannon’s diversity index, Faith’s Phylogenetic Di- versity, Evenness index, was not impacted by imidacloprid (figure 3.4). As well as this, imidacloprid did not significantly impact the distribution of bacterial taxa between groups. In the beta qualitative and quantitative analysis, I did not find a clear distance between bees exposed to imidacloprid compared with untreated bees, while phyloge- nies based on jacard distances exhibited a little separation among groups (control vs imidacloprid). Jaccard distance accounts for the presence or absence of each features in calculating the distance, giving equal weight to each OTU. Therefore It could be there is individual features that were different but are closely related.

These findings broadly support the work of other studies in this area linking imidicloprid with gut microbiota in honey bees, in which no significant changes in bacterial abundance in honey bee workers following imidacloprid exposure [55]. However, other studies showed differences in abundance of microbial community in forage honeybee workers exposed to neonicotinoid thiamethoxam [137]. This finding is contrary to our and Raymann et al. results which have suggested that forage type and neonicotinoid pesticides potential effect on gut microbiota. There are several possibilities for this discrepancy: (i) forager type (different landscapes) has a subtle effect on the composition of the bees gut microbiome. As previously shown food strongly modulates the gut microbiota in honey bee (Apis mellifera)[293]. As well as in bumble bee, Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 93 type of pollen and sugar contributed in community richness, community diversity, and the relative abundance of the gut bacteria of indoor-reared bumble bees [294]; (ii) Dis- crepancies between results could be due to difference in modes of toxic action between thiamethoxam and imidacloprid [54]; (iii) Genetic, environmental differences between biological replicate in our and Raymann et al. studies were done under controlled conditions while Jones et al. performed the experiment on thirty six honey bees colonies and in different environment landscape. It mean strong difference bacterial diversity may due to covariates between colonies including forage type, landscape and genetic background between groups (neonictinoid vs control) which confounds the result in gut bacterial diversity and composition rather environmental exposure [1].

Rare species

The bacterial species found in our samples lend further support to the presence of a core gut microbial community in bumble bee and honey bee [1, 259]. Our phylogenetic analysis demonstrate that imidacloprid exposure may have an impact on the species diversity. We found sequences assigned to the dominant Betaproteobacteria (Snod- grassella alvi) gut bacteria species in treated bees double that of control bees and also the number of species of alphaprotobacteria (alpha1 and alpha2) in bees exposed to imidacloprid increased. Number closely related species in bees gut may display mas- sive differences in functional gene content and perform various functions in the bees gut [1, 251, 295]. There is strong evidence that Snodgrassella alvi and alpha species support beneﬁcial roles to bees survival, pathogen defence and nutrient metabolism [296]. This shift of the bee gut microbiota may have detrimental effects on bumble bees health.

Surprisingly, we detected a novel pathogenic bacteria in bees exposed to imidacloprid called Paenibacillus larvae. This bacteria has not been isolated in bumble bees before but normally infects honey bee brood. Infection of our bees by Paenibacillus larvae is probably due to foods our bees were contaminated with this bacteria. We used a pollen Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 94 which originally collected by honey bee. There is one published paper about detection and isolation of Paenibacillus larvae in pollen [297]. Increasing global trade of bumble bees and honey bees for commercial purpose may aid the transmission of honey bee pathogens to bumble bees [298–300]. Managed bees may affect wild bees through sharing of some pathogenic diseases for example deformed wing virus (DWV) and the exotic parasite Nosema ceranae or slow bee paralysis virus (SBPV) [301, 302]. Another strong possibility is that there is interaction between a pathogen and an additional imidacloprid stresses. In recent years, there has been increasing evidence that sub lethal exposure to the neonicotinoid impair pollinator immunocompetence and increase disease susceptibility. López et al. [303] found a a synergistic interaction between cellular immune response of honey bee and Paenibacillus larvae when larvae are exposed to sub lethal doses of neonicotinoid clothianidin. Similarly, thiamethoxam and clothianidin has been increased rate of gut infection with Crithidia bombi and mortality rates mother queen in the bumble bee [304]. In other investigations, imidacloprid caused the suppression of on the immune response of hemocytes genes in both honey bee and bumble bee [305]. Interestingly, change in gene expression, methylation of hemocytes gene and oxidative function were observed on our study (see chapter2). In general, Lactobacillus induce immunity mechanism through protective effects against the oxidative stress. Forsgren et al. [306] have previously shown antagonistic interactions between species of Lactobacillus in bees gut to inhibit Paenibacillus larvae.

Lactobacillus helsingborgensis association with imidacloprid exposure

My results demonstrated a strong effect of imidacloprid on Lactobacillus helsingborgensis. The ratio of this lacto acid bacteria to other phylotypes increased in individuals bumble bees exposed to imidacloprid. This result is comparable to Drosophila Chapter 3. Gut microbiome effects of neonicotinoids in bumble bees 95 melanogaster studies, which have shown changes in abundance of intestinal Lac- tobacillus genera in ﬂies exposed to neonicotinoid and organophosphate insecticide [253, 256]. These studies also suggested that Lactobacillus is able to reduce organophosphate and neonicotinoid pesticide absorption and toxicity to Drosophila melanogaster. Because Lactobacillus strains exert probiotic properties in insects [256], a future issue for pesticide design might be to reduce the pesticides toxicity to lactobacillus it could be suggested that strains of Lactobacillus helsingborgensis changed here would be good probiotic candidate on how microbial biotransformation and detoxiﬁcation of neonicotinoid happen in bees gut.

The future questions raised by this study are:

• Does imidacloprid directly affect Lactobacillus helsingborgensis?

• Study the intriquing result that neonicotinoid exposure allows infection with Paenibacillus larvae.

3.5 Collaborative work statement

I conducted bees husbandry, treatment bees with imidacloprid, gut dissection and DNA extraction. Data (DNA sequence) was generated by BGI. I conducted all the bioinformatic analysis. Chapter 4

The effect of neonicotinoids on geotaxis and the circadian clock in Drosophila melanogaster

96 Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 97

4.1 Introduction

Circadian rhythms are daily activity patterns which control complex physiological and behavioural processes in living organisms, including animals, plants and cyanobacteria [307]. These biological rhythms are generated by endogenous clocks and cycle with a period of approximately one day. For example, leaf and ﬂower petals of some plants react to the sun and time of day. This rhythmic behaviour in plants regulates many biological processes, the main points being plant physiology, growth and production [308]. On the other hand, open and closing ﬂowers mirror pollinators foraging circadian rhythms [309]. This relationship is energetically favourable for bees to be most active when their food source will be ready [309].

The circadian clocks play essential role in many physiological processes, one of which is sleep. sleep is not only important for health and functions but also for their behavioral plasticity, remarkable learning capacities, and natural plasticity in circadian rhythms [309]. Bees perform high developed rhythmic activities inside and outside the nest [310, 311]. Forage workers exhibit peaks of activity during morning with increases in activity typically occurring slightly before (anticipating) the lights-on [312]. Circadian clocks enable forager bees to use sun navigation and orientating visits to ﬁnd food source and ﬂowers [309, 313]. Furthermore, walking, sitting and climbing on combs inside the nest are other important circadian behaviour [314]. Honey bees perform very sophisticated waggle dances inside the nest to recruit other foragers to the foraging site [314]. However, homing, orientation and foraging in both honey bees and bumble bee have been disrupted by insecticides and consequent failure to return to their home colony, and other abnormal daily activity [121, 315].

Despite the wealth of knowledge of the neural and molecular bases of circadian clocks in bees [309, 312], much less is know about free-running sleep period and locomotor activity rythms in bees. This is because there is no standard method to measure evening and morning activity. In Drosophila, the existence of a simple, affordable and Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 98 commercially available system for measurements of circadian locomotor activity in 12 hours of light and 12 hours of darkness condition [316].

The fruit flies are divided in two families including Tephritidae and Drosophilidae. Drosophila melanogaster, the common fruit fly, is an important model organism in modern biology because it can be readily reared, maintained and cultured in the laboratory. Drosophila melanogaster has only four pairs of chromosomes, breeds quickly, and lays many eggs, has a short generation time and has a large number of mutant strains [317]. Several behaviours of flies are under strict circadian control including general movement, eclosion and mating. Similar to most organisms the main flight activity generally takes place at dawn and another peak occurs before sunset. It has been proven to be an excellent model for locomotor activities and climbing ability against geotaxis [316, 318, 319].

My results in methylation (chapter2), gene expression (chapter2), gut microbiota investigation (chapter3) and recent other molecular studies [320] on the effects of neonicotinoid on bees has suggested an effect on the underlying circadian rhythms. Also, there are strong evidence that gut microbial factor modulates locomotor behaviour in Drosophila [321]. Intersteingly, my gene expression analysis and Colgan et al. results showed that circadian clock genes have increased in expression as effect of neonicotinoid. Imidacloprid increases bees’ locomotor activity after 15 minute of exposure at a low concentration dose (1.25 ng/bee), whereas higher doses impairs movement. Two studies by El Hassani et al. and Aliouane et al. reported that the neonicotionoid acetamiprid at sub-lethal doses increased honey bees locomotor activity [111, 322]. An- other study reported that sub lethal doses of imidacloprid, thiamethoxam, clothianidin, dinotefuran did not signiﬁcantly affect walking, sitting and ﬂying activity [323]. From these data, I conclude that locomotion and circadian rhythms in bumble bee could be effected by neonicotinoid.

In this chapter I wish to study the effects on neonicotinoids on circadian activity. Circa- dian behaviour are generally studied via the use of model organisms such as fruit flies, Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 99 mice and zebrafish. As a first step I will establish a behavioural model of neonicotinoid exposure using the fruit fly.

Brain structure is very similar between fruit ﬂies and bees [324]. In honey bees Apis mellifera, circadian rhythmic behaviours have a strong genetic element, and is controlled by multiple clock genes; Period, timeless and clock [325]. Gene structure and expression pattern for these genes in the honey bee are more similar to the mouse than to the fruit ﬂy Drosophila melanogaster [310, 325]. Several studies have suggested that the daily activity in Drosophila melanogaster plays a role in regulating the response to xenobiotics [326, 327]. So that, it could be Drosophila melanogaster offers multiple ad- vantages for the investigation effect of pesticides toxicity in bees. Many studies have found that both species are profoundly affected by insecticides in terms of toxicity. It is still not clear whether bees circadian clocks are affected by neonicotinoids or not and there is no standard method for bees circadian studies as with Drosophila melanogaster module.

The aim of this study is to investigate the effect of the neonicotinoid imidacloprid on circadian rhythms and the study will be examined by using Canton-S wild-type and M1217 strains of Drosophila melanogaster. Spontaneous physical locomotor activity was quantiﬁed during night and day periods using a monitor system with an infrared beam. Climbing ability against geotaxis was assessed using a counter-current apparatus. In addiction, this study intends to determine the level of toxicity for four different concentration of imidacloprid through 14 days of monitoring. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 100

4.2 Materials and Methods

To assess the effects of imidacloprid on locomotor activity and negative geotaxis climbing ability, Canton-S wild-type and M1217 strains of Drosophila melanogaster were selected in all experiments. They were provided by Dr Ezio Rosato, lab 124, Department of Genetics and Genome Biology, University of Leicester. The Canton-S stock was established for the ﬁrst time by Stern and Schaeffer[328]. It has a low mutation rate and it is one of the most widely used wild-type strains in genetics studies. The M1217 strain has been collected more recently by Department of Genetic and Genome Biology, Le- icester university from Taranto, Southern Italy in 2010.

Flies were maintained on standard molasses medium (containing; dry yeast, cornmeal, agar, sucrose, water and methyl 4-hydroxybenzoate) at constant 25 °C and kept in 12 hour light:dark condition.

Foods were prepared containing imidacloprid with different concentrations (2.5 ppb, 5 ppb, 7.5 ppb, 10 ppb and 20 ppb) and for a control condition distilled water was mixed with food instead of imidacloprid 5.1. Then, 1.2 ml of food was added to plastic vials lined with cotton and kept in a cold room at -4 °C.

4.2.1 Negative geotaxis assay

I used a counter current approach to assay negative geotaxis behaviour (climbing ability) in fruit ﬂies. This assay was developed initially for phototaxis experiments [319]. The apparatus is made up of 7 connected vials separated by a partition in the middle.

10 adult files (5 male and 5 female) aged between 3 - 5 days old were transferred into vials containing either control or imidacoprid food and left for 2 days allowing time to lay eggs. In order to obtain enough experimental units, fly pushing was repeated in new vial food for 3 days. After flies were born (3 days), they were transferred into clean empty vials and briefly anaesthetized with ice. Then 30 flies (15 male and 15 Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 101 female) were collected and placed in the first tube of counter current assay and left for 15 minutes to allow flies to wake up. After that, 3 sharp taps were given on a rubber mat to knock them to the bottom. They were given 20 seconds to climb into the upper tube. After 20 seconds, the flies that had climbed were moved over to the bottom of tube 2. They were again tapped to the bottom, allowed to climb for 20 seconds, and moved to the bottom of tube 3. This procedure was repeated a total of 5 times, until tube 6 was reached. At the end of the experiment, the number of flies in each tube were counted. Five replications for each concentration were conducted, using new flies for each repeat.

Statistical analysis The probability of a fly’s climbing at each trial (the climbing index) was calculated according to Kamikouchi et al. procedure which is described in footnote 4 and 5 [318]. The climbing coefficient of the final distribution is between 0 and 1. The partition coefficient Cf was calculated according to this formula:

Pn NK (k 1) k 1 − C f = Pn = (n 1) k 1 NK − =

• where n is the number of tubes (6), and and Nk is the number of the ﬂies in the kth tube6

Climbing coefficient becomes large if the flies tend to climb up, and small if they tend to stay. A two-way ANOVA was applied to see differences in climbing ability between strains, a one-way ANOVA was applied on both strains to reveal differences in climbing ability due to imidacloprid exposure and Tukey HSD (honest significant difference) was applied to compare flies climbing ability in 5 different concentration of imidaclopird. All these were implemented using R v3.4.0 [183]. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 102

4.2.2 Circadian activity

A locomotor activity monitoring system was used to quantify the circadian rhythms of flies activity during light, dark and constant darkness conditions [316]. 32 male files were collected and each fly was placed in a glass tube containing food (either imidacloprid or control) with the correct concentration of imidacloprid. Glass tubes were placed in a chamber into the activity monitor system. An infrared beam in the monitor crosses the glass tube containing the fly, which makes a recording each time the fly interrupts the beam, providing data displaying the fly’s activity over a period and allowing interpretation of their circadian rhythm. The monitor system were adjusted to be in the first condition, flies were exposed to a 12/12 hours light and dark cycle for 4 days and then kept in constant darkness condition for 10 days. The results of the locomotor activity in light, dark condition and constant darkness (5 days) were calculated and analysed using Microsoft Excel and R v3.4.0 [183]. Mean counts of recorded locomotor activity every 30 minutes in both light, dark conditions for first 4 days were calculated for each fly and mean counts for constant darkness condition were calculated for 5 days. The average activity for all 32 flies was plotted as an actogram diagram using an ImageJ plug-in ActogramJ package for Chronobiological Analyses [329]. A number of flies had single, multiple rhythmic or arrhythmic behaviour, dead flies were calculated using the CLEAN spectral analysis and auto correlation of Befly!, a custom written analysis package [316, 330].

Statistical analysis A two-way ANOVA was applied to find differences between strain (M1217 and canton-S) activity and a one-way ANOVA was applied to explore activity differences due to concentration of imidacloprid. Tukey HSD (honest significant difference) was applied to reveal the differences between variation were considered significant if P adj < 0.05 . All these analysis were implemented using R v3.4.0 [183]. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 103

4.3 Results

4.3.1 Negative geotaxis climbing analysis

A two-way ANOVA revealed that M1217 strain has a strong negative geotaxis phenotype in climbing ability compared to the Canton-S strain (F 1-39 = 1295.685, p value < 2e-16) (ﬁgure 4.1). Further analysis (one-way ANOVA) showed that there were signiﬁ- cant differences in climbing ability in Canton-S strain under different concentrations of imidacloprid (F 4-20 = 6.445, p value < 0.00168) however imidacloprid did not affect climbing activity of the M127 strain (F 4-20 = 0.989, p value < 0.437).

A TukeyHSD test was performed in order to obtain all pairwise comparisons of interaction effects between imidacloprid doses on climbing ability of Canton-S strain. A significant difference (p adj < 0.0007) was observed between 20 ppb and 2.5 ppb. The difference between 2.5 ppb and 0 ppb of imidacloprid was significant at the level p adj = 0.0286561. A near-significant trend (p adj =0.0574044) was observed between 5 pbb and 20 pbb of imidacloprid. These results suggest a significant decrease in activity in flies exposed to 20 ppb imidacloprid, whilst also indicating an increase in climbing ability when exposed to low dose (2.5 ppb) of imidacloprid. However, there were no significant differences in climbing ability against geotaxis between flies under the following concentrations of imidacloprid (10 ppb and 0 ppb, p adj = 0.79; 2.5 ppb and 10 ppb, p adj = 0.68; 20 ppb and 10 ppb, p adj = 0.24; 5 ppb and 10 ppb, p adj = 0.082; 5 ppb and 2.5 ppb, p adj = 0.99). Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 104

0.8

●

0.6

Strain

Canton−S M1217 Climbing ability

● 0.4

●

0.2

0 2.5 5 10 20 Concentration of Imidracloprid (ng/ml)

FIGURE 4.1: The plot show the effect of imidacloprid on climbing ability (negative geotaxis) of M1217 (blue plots) and Canton-S (red plots) strains. X axis, concentration of imidacloprid and Y axis, the climbing coefficient the final distribution in the six trails of the counter-current apparatus. The line inside the box indicates the median, and the bottom and top lines represent the first and third quartiles (the 25th and 75th percentiles).

4.3.2 Circadian activity

To assess the activity of flies at five imidacloprid concentrations (0ppb, 2.5ppb, 5ppb and 10ppb and 20 ppb), the average activity for each fly (32 flies) at each concentration were calculated for the following conditions; 4 days light condition (08:30 am to 08:30 pm), 4 days dark condition (08:30 pm to 08:30 am) and 5 days full darkness condition. The results of a two-way ANOVA analysis showed significant differences between strains (Canton-S and M1217) in activity during light condition (F1-8 = 183.342, P value < 2e-16 ), dark condition (F1-8 = 102.16, p value < 2e-16 ) and constant darkness condition (F1-8 = 205.672, p value < 2e-16). See figures (4.2) 12h light condition, (4.3) 12h dark condition, (4.4) 24h constant darkness condition. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 105

Activity in light condition A one-way ANOVA analysis showed, ﬁles in M1217 strain have been affected by imidacloprid (F 4-155 = 2.893, p value < 0.0241). There was a large shift (p adj = 0.0173790) in activity between ﬁles fed on food which did not contain imidacloprid and food contain 20 ng/ml of imidacloprid.

As shown in figure (4.2) an increase in activity happened when flies were exposed to high dose (20 ppb) of imidacloprid. While, the mean of activity was not affected in comparison between following concentration of imidacloprid (2.5 ppb and 10 ppb, p adj < 0.99; 20 ppb and 10 ppb, p adj < 0.77; 5 ppb and 10 ppb, p adj < 076, 0 ppb and 10 ppb, p adj < 0.28; 20 ppb and 2.5 ppb, p adj < 0.73; 5 ppb and 2.5 ppb, p adj < 0.80; 0 ppb and 2.5 pbb, p adj < 0.32; 5 ppb and 20 ppb, p adj < 0.13; 0 ppb and 5 ppb, p adj < 0.93). Also, we did not observe significant diffidence in canton-S files activity between any concentration of imidacloprid (F 4-155 = 1.672, p value < 0.159).

●

● 20 Strain ● Canton−S M1217 Activity in 12h light condition

0 ●

0 2.5 5 10 20 Concentration of Imidracloprid (ng/ml)

FIGURE 4.2: The plot shows activity in 12h light condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the ﬁrst and third quartiles (the 25th and 75th percentiles). Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 106

Activity in 12h dark condition A one-way ANOVA analysis revealed that imidacloprid had a concentration effect on ﬂies (M1217 strain) activity during 12 h dark condition (ﬁgure 4.3; (F 4-155 = 6.773, p value < 4.74e-05) while canton-S strain activity was roughly similar at all levels of imidacloprid concentration (F 4-155 = 2.086, p vale < = 0.0852).

TukeyHSD statistical tests revealed a significant differences in M1217 strain activity between flies had following concentration of imidacloprid (5 ppb and 10 ppb, p adj = 0.013; 0 ppb and 10 pbb, p adj = 0.0007; 0 pbb and 2.5 ppb, p adj = 0.02; 5 ppb and 20 ppb, p adj = 0.022 and 0 ppb and 20 ppb, p adj = 0.001) while there were no significant difference activity in comparison between 2.5 ppb and 10 ppb, p adj = 0.86; 20 ppb and 10 ppb, padj = 0.99; 5 ppb and 10 ppb, p adj = 0.92; 5 ppb and 2.5 ppb, p adj = 0.17 and 0-5 pbb, p adj = 0.91). As shown in figure (4.3), there were an increase in activity with increasing imidacloprid concentration except flies exposed to 5 ppb of imidacloprid were roughly the same as files with no exposure.

●

30 ●

●

20 ● ●

● Strain

Canton−S M1217

Activity in 12h dark condition ● 10 ● ●

● ● ● ● ● ● ● ● ●

●

● ● ●

0 2.5 5 10 20 Concentration of Imidracloprid (ng/ml)

FIGURE 4.3: The plot shows activity in 12h dark condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the ﬁrst and third quartiles (the 25th and 75th percentiles). Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 107

Activity in constant darkness condition A one-way ANOVA analysis showed no significant difference in M1217 and Canton-S strains activity at all levels of imidacloprid concentration (Canston-S strain; F 4-155 = 2.086, P value = 0.0852), (M1217 strain; F4- 155 = 1.584, p value < 0.181). As shown in figure (4.4), there is a similarity in mean activity of both flies strain in 5 days full darkness condition. But M1217 strain showed an increase in the mean activity at concentration 2.5 and 10 ppb but all these did not statistically significant.

● 40

● ● 30 ●

●

● ● Strain

● Canton−S 20 M1217 Activity in constant darkness

0 ● ●

0 2.5 5 10 20 Concentration of Imidracloprid (ng/ml)

FIGURE 4.4: The plot shows activity in 5 days constant darkness condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the ﬁrst and third quartiles (the 25th and 75th percentiles).

4.3.3 Locomotor Activity Rhythms

I found that both strains synchronized to light and dark condition cycles. As shown in figure (4.6), during the first 4 days, there were a bimodal activity pattern in both strains consisting of morning peaks (08:30 am) and evening peaks (08:30 pm). Flies in Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 108 both strains showed little activity during mid-day and mid-night while flies Canton- S strain have strong rest activity rhythms in mid-day and mid-night compared with M1217 strain (figure 4.6). Flies in 24h constant darkness showed and in all concentrations showed a normal free-running period except canton-S strain in condition 20 ppb showed a change periodicity slightly longer period) compared with other condition of same strain. As illustrated in figure (4.5) free running period of Canton-S in first 5 days of continous darkness slightly increased with increased concentration of imdiacloprid. None of these differences were statistically significant. Furthermore, as shows in table (4.1 , STD rows), there were linear associations between imidacloprid concentration with standard deviation variation of periodicity in both strains after 9 days exposure to imdicaloprid. All concentrations showed the same lethal toxicity in constant darkness in day 1 to 5 but at constant darkness day 6 to day 10, the number of dead of flies (M1217) in 20 ppb was about twice that of other concentrations.

●

● ●

●

● ● ●

● ● 28 ●

●

● ●

● ● ● strain

Canton_S ● M

period of the rhythm 24

● ●

●

● 20

●

0 2.5 5 10 20 Concentration of Imidracloprid (ng/ml)

FIGURE 4.5: The plot shows free running period in constant darkness condition, blue plots represent M1217 strain and red plots represent Canton-S strain of Drosophila melanogaster. The line inside the box indicates the median, and the bottom and top lines represent the ﬁrst and third quartiles (the 25th and 75th percentiles). DD 1-5 DD 6-10 Conc. 0 2.5 5 10 20 0 2.5 5 10 20

Canton-S N. ahr 7 4 6 5 2 10 5 8 9 8 N. dead 5 3 5 3 9 15 14 16 12 13 N. CR 0 0 5 4 2 2 4 3 5 1 N SR 20 25 16 20 19 5 9 5 6 10 period 24.27 23.83 24.61 24.9 24.65 23.12 25.54 23.27 24.47 24.67 STD 1.17 1.24 1.66 1.9 2.19 2.5 2.6 0.67 2.63 2

M1217 N. ahr 1 2 1 0 6 16 16 11 19 5 N. dead 1 1 1 1 0 11 9 14 2 24 N. CR 3 4 2 0 2 0 0 1 1 0 N SR 27 25 28 31 24 5 7 6 10 3 period 24.73 24.69 24.9 24.61 25.07 24.46 25.09 23.93 24.66 24.02 STD 1.01 0.74 1.8 1.08 2.1 1.3 1.58 0.57 3.17 0.81

TABLE 4.1: illustrating analysis of locomotor activity data in full darkness condition from day 1 to 5 (DD1- 5) and day 6 to 10 (DD6-10). N.ahr; number files did not perform stable rhythm; N.dead; number of files dead; N.CR; number of flies performed more than one thyme during 24 h; period , periodicity of rhythm locomotor activity and STD, standard division of amount of variation in periodicity between flies in same condition 0 ppb 2.5 ppb 5 ppb 10 ppb 20 ppb

Canton-S 110

M127

FIGURE 4.6: A double-plotted actogram showing average locomotor activity for 32 files in both strains of Drosophila melanogaster. X axis indicate time of day 24 h (08:30 am to 08:30 pm); Y axis indicates amount activity in every 30 minute. Each row of actogram showing showing 48 activity, each day activity showing twice belong each other except first day. A first four rows showing 4 days activity in 24 h (12 light and 12 dark) and rest rows are average activity in full constant darkness. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 111

4.4 Discussion

Here the results showed signiﬁcant difference between M1217 and Canton-S wild type strains of Drosophila melanogaster in climbing ability against geotaxis and locomotor activity in light, dark and constant darkness condition. In principle, the differences between these strains could be due to genetic variation, epigenetic or inbreeding, or a combination of these factors. M1217 strain was more recently collected from Taranto, Southern Italy in 2010 while Canton-S has been inbred and maintained in the lab since the 1960s. The M1217 strain is likely to be under selection in natural populations and therefore the M1217 strain will have higher genetic variation than the Canton-S strain.

Inbreeding Canton-S strain makes flies more likely to have genetic disorders [331]. It does affect flies ability to survive or adapt. This finding broadly agrees with the work of other studies in this area linking genetic variability with activity, response to disease and environmental stress [332–334].

In comparing each strain’s susceptibility toward imidacloprid, low dose (2.5 ppb) increased climbing ability of Canton-S flies against geotaxis while the highest doses (20 bbp) negatively affected climbing ability. Imidacloprid at all concentration levels did not effect climbing ability of M1217 strain. These results were contrary to the locomotor activity assay. Activity of M1217 strain in light and dark condition were significantly affected by imidacloprid while Canton-S flies activities were not effected by any doses of imidaloprid.

Of course, these contradictions between strains in circadian activity respond to imidacloprid may simply indicate that imidacloprid can effect the circadian rhythms in multiple directions but the effects depend on strain. Circadian behaviours effects of neonicotinoids on these strains of Drosophila melanogaster has not previously been studied. Therefore, it is difﬁcult to suggest the accurate reason for these results. It might be both strains have their own different resistance genes to imidaclorprid in the Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 112 line previously discovered differences between strains of Drosophila melanogaster in resistance to insecticides [335].

Furthermore, the current study found that imidacloprid at 2.5, 5, 10 and 20 ng/ml concentration for 9 days exposure did not effect on survival of both strains of Drosophila melanogaster. But after extra five days (14 days) in constant darkness most M1217 flies in condition 20 pbb were dead (24 out 32 flies) whereas Canton-S flies were still resistant. It could be due to that Canton-S strain never exposed to imidacloprid because it has been inbred in the laboratory more than 60 years while neonicotinoid were only developed the 1980’s. M1217 flies were collected 8 years ago therefore might have been previously exposed to neonicotinoids.

The cytochrome P450 super family enzyme is responsible for the tolerance of Drosophila melanogaster to neonictinoids [336, 337]. In Drosophila melanogaster an over expression of cytochrome gene correlates with increased survival in resistance to pesticide, while cytochrome knock-down dramatically reduce survival [225]. My gene expression analysis (chapter 2) indicated that 6 cytochrome genes in non reproductive workers of Bombus terrestris were imidacloprid responsive and interestingly 4 of them were down regulated.

Surprisingly an increase in activity was absent when flies were fed on foods contain 2.5, 5 and 10 ng/ml of imidacloprid. This result is comparable with previous studies on the increase bees activity as result neonicotionoid exposure [111, 322]. Other most obvious finding to emerge from the data analysis is a decrease in activity with increase in imidacloprid to 20 pbb. These results corroborate the findings of a great deal of the previous work in showing that acute exposure thiamethoxam alters honey bees activity, motor functions, and movement to light [124] while chronic exposure significantly increase forager movement to light. Bees and fruit flies drank more when food contain low dose of neonicotinoid [238]. More consumption of foods may give bees and flies more power. Abnormal bees activities could impair bumble bee health by harming Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 113 worker locomotion and potentially alter division of labour if workers going to foraging or remain inside the nest.

As shown in table (4.1), there is a similarity between flies in all conditions regarding periodicity rhythmic activity. But the variation in periodicity in 1-5 full darkness condition between Canton-S flies in same condition linearly increased with increased imidalcoprid concentration and Standard deviation were doubled in M1217 between condition 0 and 20. Several studies have suggested that molecular mechanisms of cycles locomotor activity in Drosophila melanogaster play a role in regulating metabolism and detoxification of drugs [326, 338]. It is therefore likely that increase concentration of imidacloprid increased metabolic process which would be reflected in periodicities of the test organisms during the course of exposure.

In summary, in the present study, I investigated the effect of imidacloprid on circadian rhythms using Drosophila melanogaster model. The experiments confirmed that all doses of imidacloprid (0 to 20 ppb) used in the study, are non lethal doses 9 days exposure. Imidacloprid has an effect on climbing ability and locomotor activity in dark conditions. Surprisingly, low dose increased flies activity. Another finding is that there was a discrepancy in susceptibility between strains to imidacloprid and the direction of effect depended on Drosophila melanogaster strain. M1217 strain was more sensitive to imidacloprid than Canton-S strain in activity in during light and dark condition and vice versa in locomotor activities. Taken together, these findings suggest a role of imidacloprid in affecting Drosophila melanogaster circadian clocks. Further, molecular studies on candidate clock genes ( such as Period, Cryptochrome and Clock) are important to be carried out in order to validate the effect of imidacloprid on bumble bee and honey bee. Chapter 4. Behaviours effects of neonicotinoids in the Drosophila model 114

4.5 Collaborative work statement

All the work in this study were conducted by myself and I got assistance in running locomotor activity system experiments and data generating process from Dr Ezio Rosato. Chapter 5

The effect of black carbon on bacterial community of B. terrestris workers

115 Chapter 5. Black_Carbon affects on gut microbiota 116

5.1 Introduction

Environmental pollutants like particulate matter pose a clear threat to both humans and wildlife in polluted areas [339–341]. The main sources of particulate matter are both man-made (e.g., diesel engines) and natural processes (e.g., wildfire) [342]. Par- ticulate matter is classified into two groups; primary particles are released directly into the atmosphere through burning of fuels for transport, industrial, commercial and do- mestic purposes. Secondary particles are formed through atmospheric chemical reactions. Primary particles include black carbon, sodium chloride, and trace metals. Sec- ondary particles include sulphates and nitrates, such as ammonium sulphate and ammonium nitrate formed by the oxidation of sulphur dioxide (SO2), nitrogen oxides (NO and NO2), and ammonia [339, 343]. Particles range from several micrometers (PM10 between 2.5 and 10 microns (micrometers)) to a few microns (micrometers) (PM 2.5 and 0.1) in diameter, a human hair is about 60 micron in diameter. Particular sizes of particles can remain suspended in the atmosphere for about one week [344]. Fine (2.5 PM) and ultra fine (0.1 PM) particles move easily through the air, which can travel over 100 km, thereby species developing in polluted area may be potentially exposed [343].

Fine particulate matter has the potential to cause health problems [345]. For example, studies showed that black carbon is easily inhaled, so penetrates the respiratory system, cross the blood barrier and then are distributed to most organs. Finally it could be play a role in toxicity, pathogenesis and ﬁnally lead to death. Data from several studies found that air pollution cause seven million deaths each year, 3.7 million deaths are caused by exposure to ﬁne particulate matter [339, 345].

In the UK, the air quality expert group was set up in 2002, the main purpose for this group is to give assessments about air quality; including measurement the composition and current and future concentrations of particulate matter across the UK. The composition of fine particulate matter is varied and depends on emissions, weather Chapter 5. Black_Carbon affects on gut microbiota 117 conditions, local and regional contributions, and temporal variations [346]. Black carbon is an essential component of fine particulate matter (2.5 PM, smaller than 2.5 microns) and it is released to the atmosphere through the incomplete combustion of fossil fuels, biofuels, and biomass [12]. Extensive research during 2009 – 2016 has shown that the concentration of black carbon is 15 PM2.5/µgm-3 across the UK. The concentration increase to near 20 PM2.5/µgm-3 in winter whereas in summer decreases to 10 PM2.5/µgm-3 at urban background sites [11]. Another study in 2014, levels of black carbon ranged from 1 to 7mg/m3 across the UK [347]. The Air Quality Expert Group suggest that walking and cycling in close proximity to road traffic increases the possibility of black carbon exposure. Many studies have shown that inhalation of black carbon can causes serious health problems including inflammation, oxidative stress and alteration in the respiratory and cardiovascular disease, cancer and even significantly impacting the immune response by impairing macrophage function [339, 348, 349].

Particulate matter and wildlife A study looked broadly at the impacts of particulate matter on bird health advocated that birds suffer the same respiratory problems as humans when exposed to air pollution. Bees foraging and pollination efﬁciency decreases with increasing air pollution[14]. Fine particulate matter were detected in bees bodies after foraging performance [15]. This indicates that bees health, like birds and humans may be effected by the ﬁne particulate matter. Till now nothing is known about impact of particular matter on bees health, practically on gut microbiota. In this study, we are looking at the possible effects of particulate matter on gut bacteria in the worker bumble bees.

Currently, a group in lab 121, Department of Genetic and Genome Biology are looking at the effects of black carbon on microbiota in mice module and bumble bees. Recently Hussey et al. showed that black carbon signiﬁcantly affects bacterial colonisation and bioﬁlm formation [66]. Chapter 5. Black_Carbon affects on gut microbiota 118

As I described in (see chapter 3), bees guts contain many bacteria species and these microbes are important for bees health in many ways [251, 350, 351]. However, bees gut microbiota are very sensitive to various pollutants [137, 273, 352, 353].

In this chapter, I hypothesise that black carbon changes bee gut microbiota in a way that negatively impacts bumble bee health. We explore this possible differences between workers exposed to black carbon and unexposed, using culture based methods, absolute quantiﬁcation method by performing real time quantitative polymerase chain reaction (RT-qPCR) approach and next generation sequencing (metagenomics) of 16S rRNA gene regions (V1 - v2) using Ion Torrent sequencing technology. Chapter 5. Black_Carbon affects on gut microbiota 119

5.2 Materials and methods

5.2.1 Bee Husbandry and experimental design

Four colonies of Bombus terrestris audax were reared as previously described in section 2.2.1. The project was conducted in two rounds. One colony was used for culture based methods and qPCR approach. Three colonies were for next generation sequencing of 16s rRNA gene for metagenomic analysis.

5.2.2 Black Carbon preparation

Stock black carbon solution was provided by Dr. Julie Morrissey, lab 121, Department of Genetics and Genome Biology, University of Leicester. which was purchased from Sigma-Aldrich (UK) under product number 699632. This was provided as a powder with a size distribution of <500 nm, with <500 ppm trace metals, and a weight of 12.01 g/mol and the concentration of black carbon in the stock solution was 10 mg/ml [66]. Lab 121 use dosages from 100 µg/ml to 2 mg/ml in testing effect of black carbon on gut microbiota in mice. However I exposed the bees to black carbon only once, rather than chronically as in the mice. At their recommendation, I diluted to 7 mg/ml of apiary syrup by mixing black carbon with sugar water and the dilution was carried out in comparing between mice body mass and bumble bee body mass.

First experiment

5.2.3 Black carbon exposure and Faeces collection

30 callow workers (less than 12h old) were collected from one colony. 15 bees were treated with black carbon and 15 were control bees; essentially 3 replicate groups per each treatment group, 5 bees in each replicate set. When bees reached 3 days old, Chapter 5. Black_Carbon affects on gut microbiota 120 each bee in treated groups was fed separately with 15 µl of black carbon (concentration 7 µg/ml) while control groups were fed with sugar water. This was performed by pipetting treatment solution into an individual sterile plastic vial. After two days of exposure, bees were transferred to perspex box in plastic vials with air-holed lids (faeces collection was carried out within 2 days). About 100 µl of faeces were collected from each group (3x control and 3x black carbon). 20 µl of faeces was cultured in agar plates. The rest of faeces (80 µl) were stored in a freezer at -20 °C for qPCR assay.

5.2.4 Preparation culture media

BHI (Brain-Heart Infusion Agar) and MRS Agar were obtained from Lab 121, Depart- ment of Genetics, University Leicester. Media plates were prepared according to manufacturers protocol.

5.2.5 Bacterial cultivation

20 µl of faeces was diluted in 80 Phosphate-buffered saline (PBS). 10 µl of faeces solution was added to the centre of BHI and MRS plates. The faeces were spread out over the surface of agar using a sterile cotton swab. The plates were incubated at 37°C with 5% CO2 and 35°C with 10% CO2. There were two replicate culture plates for each experimental groups.

5.2.6 Classiﬁcation bacterial colony

BHI and MRS plates cultures were incubated for 24h and 72h respectively. Number of colony (Colony Forming Unit CFU) in each plate was calculated. I used the median of the number of colonies between two plates as the dependent variable and logged the data to make the residuals normally distributed. The analysis of variance (ANOVA) was Chapter 5. Black_Carbon affects on gut microbiota 121 done on the logged data to check for statistical differences between control and black carbon group using cut-off p value < 0.05.

Different bacterial colonies were identiﬁed according to their morphology, size, surface, texture, color, elevation and margin or edge of the colonies (Bergey’s Manual of Systematic Bacteriology). Colonies showing differences in morphology were selected and observations performed under a dissecting microscope. To obtain pure culture media, a single colony was taken by a loop and then streaked out on fresh agar (BHI media) using the streak plate technique. All pure culture plates were sorted out and named by speciﬁc code. Again a single bacterial colony was selected from pure culture for DNA extraction and rest of bacterial colonies were suspended in 20% of TBS Glycerol and stored at -80 °C for future work.

5.2.7 DNA extraction

A single colony was suspended in 200 µl of PBS. The bacterial suspension was centrifuged at 13000 rpm for 5 minutes. The supernatant was removed and 100 µl of distilled water was added to the precipitate. In a thermocycler machine, the solution was heated for 5 minute at 98° and then centrifugation was carried out again at 13000 rpm for 5 minutes. The supernatant was placed in a sterile tube. The quality and concentration of DNA were assessed using a NanoDrop. The DNA solution were stored at - 20 °C.

5.2.8 Ampliﬁcation of 16s rRNA gene

The Polymerase Chain Reaction (PCR) was performed to amplify the Domain B hypervariable regions (V4-V5 region) of the 16S rRNA gene. All the regents were prepared and then mixed in a PCR tube: 0.5 µl forward primer “AGAGTTTGATCCTGGCTCAG” conc. (10 NM), 0.5 µl reverse Primer “ACGGCTACCTTGTTACGACTT” conc. (10NM), 2.5 of 10 Kapa Buffer, 0.5 µl dNTP con (10 mM), 0.1 µl Kapa Taq polymerase (5 u/ µl), Chapter 5. Black_Carbon affects on gut microbiota 122

1 µl bacterial DNA and 19.9 µl dH2O. One solution was prepared with dH2O (without bacterial DNA) as a negative control. Finally, the samples were run in the thermocycler machine. Steps are listed in table (5.1), denaturation, annealing and extension were carried out on 30 cycles. The PCR product were puriﬁed using E.Z.N.A.® Cycle Pure Kit following manufacturing procedure.

No steps temperature Time

1 Activation 95 °C 3min

2 Denaturation 95 °c 30s

3 Annealing 52 °C 30s

4 Extension 72 °C 30 min

5 ﬁnal Extension 72 °C 30 min

TABLE 5.1: Hot and cool cycler steps of PCR ampliﬁcation 16s rRNA gene using Kapa Taq Polymerase

5.2.9 Agarose gel electrophoresis

Agarose gel electrophoresis was used to analyse PCR ampliﬁcation products. The gel was prepared by mixing 1g agarose in 100 mL Tris-Acetate-EDTA (TAE), 2 µl of ethid- ium bromide. After autoclaving, the gel was submerged in a gel tank to about 4 mm depth. Then, 2 µl of HyperLadder 1 were loaded to the ﬁrst well and others wells were loaded with mixture of (5µl of PCR product and 1 µl loading buffer and the last well was loaded with negative controls. The immigration was run at 100 volt for 30 min. Finally, the gel was visualized on a UV transilluminator (Fisher). The qualitative result of PCR products (positive DNA band) was captured out using an image scanner (Gene Genius - Bio Imaging System). Chapter 5. Black_Carbon affects on gut microbiota 123

5.2.10 DNA sequencing preparation

In a PCR tube, I mixed 2 µl of purified PCR product, 0.5 µl of BigDye® 3.1, 1.75 µl sequencing buffer, 0.35 µl of primer and 0.4 µl of dH2O together. After reagents were mixed by pipetting several times, 16s rRNA amplification was carried out using a ther- mocycle machine for each sample separately and then PCR products were purified using Edge Bio Performa® DTR according manufacture protocol.

Denaturation, annealing and extension were carried out on 30 cycles and the temperature for each steps are listed in table 5.2 and following primers were used:

• 16s-536F “CAGCAGCCGCGGTAATAC”

• 16S-536R “GTATTACCGCGGCTGCTGCTG”

• 16s-1050F “TGTCGTCAGCTCGTG”

• 16s-1050R “CACGAGCTGACGACA”

No steps temperature Time

1 Denature 96 °C 10s

2 Denature 96 °c 30s

3 Annealing 50 °C 10s

4 Elongation 60 °C 4 min

5 Elongation 60 °C 5 min

TABLE 5.2: showing stages of PCR process Chapter 5. Black_Carbon affects on gut microbiota 124

5.2.11 PCR product puriﬁcation

PCR products were cleaned up using a Edge Bio Performa® DTR gel ﬁltration car- tridges. First in a clean PCR tube all the following regents were mixed 5 µl PCR product, 5 µl dH2O and 1 µl of SDS at 2.2%. After that, the samples were incubated in the thermo-cycler machine at 98 °C for 5 min, then were incubated at 25 °C for 10 min. The columns were centrifuged at 3100 rpm for 3 min. Then, columns were placed in a new centrifuge tube. Finally, The tubes was centrifuged again at 1300 rpm for 3 min.

5.2.12 Data sequencing processing

Puriﬁed PCR products were sequenced by the protein nucleic acid chemistry laboratory (PNACL) at the University of Leicester. Sanger sequences were visualized using the Geneious R8 software programme. Low quality bases at the beginning and end were trimmed then the sequences coming from same samples (primers: 16s-356F - 16s-356R, 16s-1050F - 16s-1050R) were aligned against each other. The consensus sequences were extracted and DNA sequences were blasted against genbank of the National Center for Biotechnology Information (NCBI) database using the online Ba- sic Local Alignment Search Tool (BLAST) program ( https://blast.ncbi.nlm.nih. gov/Blast.cgi ).

5.2.13 Real time qPCR Quantiﬁcation

5.2.13.1 DNA extraction, 16s rRNA ampliﬁcation and puriﬁcation

The following steps were conducted to determine the size (in bp) of the 16s rRNA fragment in the faeces of bumble bees:-

1. 100 µl faeces was collected from non-experimental bees. Chapter 5. Black_Carbon affects on gut microbiota 125

2. Total DNA was extracted using a Fast® DNA SPIN Kit for Soil and the FastPrep® Instrument (MP Biomedicals, Santa Ana, CA), as per manufacturer’s instructions.

3. PCR was performed in a 96-well microtitre plate (Biometra T Gradient Ther- moblock Thermal Cycler) to amplify hypervariable regions fragments (V4-V5 region - 389 bp) of 16s rRNA gene.

4. To verify the correct ampliﬁcation size ( 389 bp) and estimate concentration of target sequence, PCR products were run on a standard agarose gel and the DNA band was visualized under a UV lamp. DNA fragments was extracted and puriﬁed from gel using a Wizard® SV Gel and PCR Clean-Up System as following:

• Get rid of all excess gel, DNA band was excised by cutting gel carefully using a sharp scalpel.

• DNA band (gel slice) was dissolved in membrane binding solution then in- cubate at 60 °C until gel slice was completely dissolved.

• DNA solution was centrifuged at 16 000 x g for 1 minute then mini column assembly tube was cleaned up by washing with ethanol then centrifuged again. To allow evaporation of any residual ethanol, assembly tube was kept at room temperature for 2 minute in a new clean microcentrifuge tube.

• 50 µl of nuclease-free water were added to the mini column then centrifuged for 1 minute and DNA solution kept in -20 °C for late use.

5. Genomic DNA were precipitated using isopropanol method to concentrate and / or purify nucleic acids.

6. DNA concentration was calculated using a Qubit 2.0 Fluorometer assay kits. Chapter 5. Black_Carbon affects on gut microbiota 126

5.2.13.2 Creating a gDNA standard curve

A standard curve is needed in quantitative real time PCR analysis. The target sequences of PCR product was total bacteria (389 bp) of 16s rRNA gene (Forward: CA- GAGCCGCGGTAATAC - Reverse: CCGTCAATTCCTTTGAGTTT). The following steps were performed to determine the mass of gDNA.

Step 1 :

DNA mass of 16s rRNA gene (target region) was identiﬁed according to this formula: mass = [n] [1.096e-21g/bp] where: n = genome size (bp) = target gene (389 bp) mass = 4.26344e-19 g. The calculation below was applied to convert the mass unit from ng to picogram units

4.26344e-19 g * 1e12/g = 4.26344e-07 pg

Step 2 :

The mass of the target gene was multiplied by the copy number of the target gene. 4.26344e-07 pg/genome * 1 copy of fragment/genome = 4.26344e-07 pg. Therefore, 4.26344e-07 pg of DNA required for one copy of the of the target gene.

Step 3 : The mass of target gene was calculated containing the copy#s (#s is copy number of target gene) of of interest, that is 300,000 to 30 copies (table 5.3). Chapter 5. Black_Carbon affects on gut microbiota 127

Copy# Mass per one fragment Mass of gDNA required (pg)

300,000 1.27e-01 30 000 1.27e-02 3000 X 4.26344e-07 pg 1.27e-03 300 1.27e-04 30 1.27e-05

TABLE 5.3: shows calculation the mass of gDNA (16s rRNA ) containing the copy #s of interest

Step 4 : To achieve the copy#s of interest in our DNA sample (5 µL of gDNA solution), mass needed (calculated in Step 3) was divided by the volume to be pipetted into each reaction (table 5.4).

Final conc Copy# Mass of gDNA needed (pg) – (pg/µl) of gDNA

300,000 1.27e-01 2.54e-02 30 000 1.27e-02 2.54e-03 3000 1.27e-03 /5 µL 2.54e-04 300 1.27e-04 2.54e-05 30 1.27e-05 2.54e-06

TABLE 5.4: shows the concentrations of gDNA needed to achieve the copy#s of interest

Step 5 : Serial dilution of the gDNA were prepared from stock solution (108 ng/µl). C1 * V1 = C2 * V2 formula was used for the dilutions.

DNA concentration from bees faeces was 0.108 µg/µl. Therefore, the concentration of stock solution, C1 = 0.108 ng/µL or 108000 pg/µL. Three times dilution was carried out for stock solution using nuclease-free water; - 3x stock solution = 0.108pg/µL

0.108 pg/µL * V1 = 100 µL * 2.54e-02 = 2.35 µL Chapter 5. Black_Carbon affects on gut microbiota 128

Volume of diluent = 100 µL – 2.42 µL = 97.65µL.

The rest of dilution were prepared to obtain samples containing 30,000, 3,000, 300, and 30 copies of 16s rRNA gene (table 5.5). l) µ l L) µ L) µ L) µ L) µ µ Final DNA gene/ 5 (pg/ of 16s rRNA Final conc of Volume of Volume of Volume ( Dilution Bees Faeces Initial conc. gDNA ( diluent ( dilution (pg/ Resulting copy

– – C1 V1 – C2 V2 –

1 - 3x stock 0.108 2.42 97.5 100 2.54e-02 300 000

2 Dilution 1 2.56e-02 10 90 100 2.54e-03 30 000

3 Dilution 2 2.56e-03 10 90 100 2.54e-04 3 000

4 Dilution 3 2.56e-04 10 90 100 2.54e-05 300

5 Dilution 4 2.56e-05 10 90 100 2.54e-06 30

TABLE 5.5: showing the calculated volumes of gDNA and diluent for all 5 dilutions

Step 6 : Stock solution of 16S rRNA gene DNA was diluted as illustrated in table 5.5. The qPCR was carried out as described in section 5.2.13.2 and each qPCR reaction were done in triplicate. Fast SYBR® Green Master Mix (Applied Biosystems) was utilised in the qPCR method as per manufacturer’s protocol. 1 µL forward and reverse primer (total bacteria V4), 12.5 µL SYBR green, 5.5 µL nuclease-free H2O were mixed in 96-well optical plates (Thermo Fisher Scientiﬁc) and sealed with MicroAmp® Optical Adhe- sive Film (Applied Biosystems). The regents were mixed using MixMate® (Eppendorf) at 550 rpm for 30 seconds and subsequently centrifuged using the Jouan B4i Multifunc- tion Centrifuge (Thermo Electron Corporation) at 2000 rpm for 1 minute.

The qPCR reactions were carried out in a LightCycler® 96 (Roche, GmbH) and the results were processed using LightCycler® 96 Software. The quality of all the qPCR reactions were assessed according to the melting and amplification curves. Mean of Ct Chapter 5. Black_Carbon affects on gut microbiota 129 values were calculated for all three replication (Ct values that were not matches and outliers among the triplicate was discarded in further analysis, I decided this upon vi- sualizing of melting curve). All data was analysed in Excel from the data obtained after qPCR run. A standard curve was generated (figure 5.1) by plotting the mean of the cycle threshold (Ct) value against log DNA concentration (copy number). The efficiency of the PCR (E) was taken into account by the equation E = 10ˆ(-1/slope). In our analysis, R2=ˆ 0.99717, coefficient of determination is a parameter which indicates that there was strong confidence in the correlation between the copy number of fragment and cycle threshold values, the amount of product doubles with each cycle [354].

FIGURE 5.1: An example of absolute quantiﬁcation using a Standard curve. liner graph generated from the Ct values of stock (300 000 copies), dilution 1 (30,000 copies), dilu- tion2 (3,000 copies), and dilution 3 (300 copies). The graph shows the mean of the cycle threshold (Ct) value along the Y axis against log DNA concentration (copy number) along the x axis. The efﬁciency of the PCR (E) was taken into account by the equation E = 1.0e-1 /slope -1. Chapter 5. Black_Carbon affects on gut microbiota 130

5.2.14 Experimental qPCR Setup

In order to confirm the presence of bacterial DNA in the six samples (3 black carbon and 3 control), PCR was used to amplify the phylum- specific (Alpha protobacteria, Be- taproteobacteria, Firmicutes, Actinobacteria and Bacteroidetes) of the 16S rRNA gene (table A.6). qPCR was performed for each sample using Fast SYBR® Green Master Mix (Applied Biosystems) with similar steps as step 5 (5.2.13.2). DNA of these phlyum were amplified and analysed as in sections (5.2.8 and 5.2.9).

The standard curve was used to calculate absolute copy number of each treatment sample (black carbon and control) using the absolute quantiﬁcation method and subsequently the copy numbers were determined in each sample.

Target Forward primer set Reverse primer set Size (bp) Ref

DomainB : DomainB: Total Bacteria 398 [355] CAGAGCCGCGGTAATAC CCGTCAATTCCTTTGAGTTT

Alpha- Eub338: Alf685: 365 [356] proteobacteria ACTCCTACGGGAGGCAGCAG TCTACGRATTTCACCYCTAC

Beta- Eub338: Bet680: 360 [356] proteobacteria ACTCCTACGGGAGGCAGCAG TCACTGCTACACGYG

Actino: Eub518: Actinobacteria 300 [357] CGCGGCCTATCAGCTTGTTG ATTACCGCGGCTGCTGG

Lgc353: Eub518: Firmicutes 180 [356] GCAGTAGGGAATCTTCCG ATTACCGCGGCTGCTGG

Cfb319: Eub518: Bacteroidetes 180 [356] GTACTGAGACACGGACCA ATTACCGCGGCTGCTGG

TABLE 5.6: showing the sequence and amplicon size of the primers used in qPCR approach to look for broad changes in bacterial species after exposure to black carbon Chapter 5. Black_Carbon affects on gut microbiota 131

5.2.15 Quantiﬁcation of total and speciﬁc gut phyla

Real Time qPCR was performed using iQ SYBR Green Supermix (Bio-Rad) as in the same approach used in section 5.2.13.2. I used the standard curve to analyse (absolute quantiﬁcation) the effect of black carbon on total bacteria.

The Pfafﬂ method ([358], formula I) was applied to calculate the relative bacterial concentration between control and black carbon groups for all speciﬁc phyla. Differences in Ct values in total bacteria between the control and black carbon groups were chosen as the calibrator. The ratio of fold change between control and black carbon was calculated using the following formula:

‘E’ refers to the ampliﬁcation efﬁciency of the target (phylum) and reference (standard curve) was calculated for each individual plate.

10 (C ttreat C t )tar get − − ctrl Etar get Ratio 10 (C ttreat C t ) = − − ctrl re f erence Ere f erence

• (Cttreat - Ctctrl) target = difference between cycle threshold between black carbon and control for the speciﬁc target phylum.

• (Cttreat-Ctctrl) reference = difference between cycle threshold between black carbon and control for the reference gene (which is difference in total bacteria between groups).

• If the ratio is below 1 this indicates that there are less bacterial cells in black carbon samples compared to controls. Chapter 5. Black_Carbon affects on gut microbiota 132

Second experiment

5.2.16 Black carbon exposure, faecal collection and DNA extraction

A total of 150 workers were collected from three colonies of bumble bees. 50 bees from each colony in which 25 bees were treated with black carbon and 25 bees with treated sugar water. We used 5 replicate groups per treatment type. Bees were transferred into Perspex boxes (micro colony) and fed ad libitum with sugar solution and bee pollen; each micro colony contained 5 bees of the same treatment type. To identify each worker in the box, bees were tagged (see section 3.2.2). There were a total of 30 groups; (3 colonies X 5 replicate X 2 condition). After 3 days in the micro colonies; faeces were collected from all treatment groups as pre exposure then after one day bees were fed with treatment solution (either black carbon or sugar solution). After two days bees were transferred in plastic vials with air-holed lids and anaesthetised on ice for 10-15 minutes. As soon as they were anaesthetised, a gentle pressure was applied to the workers abdomen and faeces were released into 1.5 mL sterile Eppendorf tubes. As soon as the collection was completed the bees were returned immediately back to their boxes. Faeces were collected twice a day from each bee into a single tube and stored at - 20 °C for later use. Because not all bees produced enough faeces the ﬁrst and/or second collection and in order to obtain a substantial amount of faecal material to be used for DNA extraction (50 µL). The faeces from the same micro colony were pooled according to colony, treatment type and time point. DNA was isolated from a total of 24 faeces samples using the FastDNA® SPIN Kit for Soil and the FastPrep® Instrument (MP Biomedicals, Santa Ana, CA), as per manufacturer’s instructions and stored at -20 °C for later use. Chapter 5. Black_Carbon affects on gut microbiota 133

5.2.17 16s rRNA gene sequencing

First, DNA concentration was checked using NanoDrop™ 2000/2000c Spectropho- tometers (Thermo Scientific) and normalised for all samples to 20 ng per sample. PCR was performed for next generation sequencing targeting of 16S rRNA gene (HVR V1-V2) using Ion Torrent linker primers with barcode sequences, forward primer 8F “AGAGTTTGATCCTGGCTCAG” and reverse primer 334R “TGCCTCCCGTAG- GAGTCTG” and with the different barcode sequences were mixed for each sample, all were obtained from Sigma® Life Science.. The following cycling conditions were applied in PCR amplification: initial denaturation at 98 °C for 30 seconds; then 25 cycles of denaturation at 98 °C for 10 seconds, annealing at 50 °C for 30 seconds and extension at 72 °C for 20 seconds, with a final extension step at 72 °C for 2 minutes.

Second, the PCR product ( 400 bp) was analysed using agarose or polyacrylamide gel electrophoresis in Tris-Borate-EDTA (TBE) buffer. Then, DNA was extracted and pu- riﬁed from the gel slice as mentioned in section (5.2.13.1). Concentration and purity of the gDNA were checked using a Agilent 2100 Bioanalyzer (Agilent Technologies, UK) and different dilutions: 1:5, 1:25, 1:50, 1:75 and 1:100 were prepared in order to determine the most suitable molarity for Ion Torrent sequencing.

Next-generation sequencing of gDNA was performed in Leicester university - genetic and genome biology department - lab 19. the Ion PGM™ System (Life Technologies) with the use of Ion PGM™ Hi-Q™ View Sequencing Kit and Ion PGM™ Hi-Q™ View OT2 Kit; and the Ion PGM™ Chip enabling 400 bp reads. 100 pM of the library was used for sequencing.

5.2.18 Bioinformatics analysis

All analysis were performed on QIIME 2 Core v2018.6 pipeline [287]. First, Ion torrent data was imported into a QIIME 2 artifact using qiime tools import plugin then multi- plexed reads with the barcodes in the sequence reads were demultiplexed according to Chapter 5. Black_Carbon affects on gut microbiota 134 the barcode sequence which associated with each sample. qiime cutadapt trim-single function was applied to trim barcode sequences for each sample. Trimmed reads were visualized to determine how many sequences in each sample, and to calculate a summary of the distribution of sequence qualities at each position in our sequence data. Sequence quality control and feature table was constructed (OTU or BIOM table) with dada2 denoise-single plugin in qiime2. This was implemented to remove the ﬁrst 5 bp nucleotide and the last 75 bp in all samples and error sequences and base pairs were removed. I generated a tree for phylogenetic diversity analyses (Alpha and beta diversity analysis), taxonomic analysis and differential abundance analysis with gneiss were performed based on differences between bees exposed to black carbon and untreated bees (see chapter3 - section 3.2.4). Chapter 5. Black_Carbon affects on gut microbiota 135

5.3 Results

5.3.1 Bacterial isolation and identiﬁcation

Colonies, from all different bee faeces samples, grown after 24 hours on BHI agar plates were white, smooth and varied little in size. Optically there were no differences in bacterial growth between plates in conditions of 5% and 10% of CO2. There were only two small colonies on MRS media (1 plate) in black carbon group which failed to be re- cultured again. This ﬁnding is contrary to previous studies which have found that bees gut bacteria grows on the MRS agar media [359].

I cultured and purified bacterial colonies which showed differences in morphology on new BHI agar plate successfully. I isolated 19 bacterial colonies with observable differences in morphology. I amplified and sequenced fragments of 16s rRNA by using 16s-536R, 16s-1050F and 16s-1050R primers in 13 of these (the other six failed to sequence). 11 strains belonged to phylum Firmicutes, 2 strains assigned under the phylum Gammaprotobacteria, one genus of phylum Acintobacteria and one isolate was identified under unknown bacteria (see table 5.7). Chapter 5. Black_Carbon affects on gut microbiota 136

Phylum Best match Seq bp Accession number

Firmicutes Bacillus cereus 1505 LN890264.1 Firmicutes Bacillus 1505 LN890264.1 Firmicutes Staphylococcus warneri 1504 NR-102499.1 Firmicutes Lysinibacillus fusiformis 1507 HG421013.1 Firmicutes Lysinibacillus fusiformis 1503 HG421013.1 Firmicutes Oceanobacillus sojae 1517 NR-112845.1 Firmicutes Oceanobacillus sojae 1524 NR-112845.1 Firmicutes Staphylococcus 1505 LN890264.1 Firmicutes Staphylococcus 1505 LN890264.1 Actinobacteria Micrococcus 1505 LN890264.1 Gammaproteobacteria Pseudomonas sp 1486 EF157292.1 Gammaproteobacteria Gilliamelia apicola 1489 EF157292.1 Unknown uncultured bacterium 1505 LN890264.1

TABLE 5.7: Shows bacterial species which were isolated from faeces B. terrestris colonies On BHI media. Blasting query cover For species in NCBI database were 100 % .

5.3.2 CFU analysis

We performed counts of CFU that were able to form visible colonies within 1 days of inoculation on BHI agar plates at 10% and 5% of CO2 37 °C. After log transformation of median counts of CFU, model residuals plots showed normal distribution of data across all samples. The ANOVA showed that median count of CFU were not different

(F1-8 = 1.464; p value = 0.2608 ) between CO2 conditions. So the median counts (CFU) of plates under different CO2 conditions were combined together. The ANOVA (one way) test showed that median count of CFU were different (F 1-8 = 20.981; p value < 0.0018 ** ) between bees treated with black carbon and untreated bees. AS shown in graph (5.2), there was an increase in bacterial cells in bees exposed to black carbon compared with control bees. Chapter 5. Black_Carbon affects on gut microbiota 137

200

150

100 Number of colonies

Control Black Carbon Treatment

FIGURE 5.2: Shows median count of colony forming unit in faeces of bees exposed to black carbon and control bees. X axis represents treatment condition and Y axis represent number of bacterial colonies

5.3.3 Quantiﬁcation of bacterial gut community

Absolute quantification of the total bacteria community in fecal samples was quantified using the Domain B V4-V5 primer set with the qPCR approach. Results showed a significant differences in total bacteria between control bees and treated bees (black carbon). Bees which were treated with black carbon showed a greater bacterial population ( average copy number = 5863132.766; stdev = 4255871.062) in their faeces, whereas bees were fed with sugar water showed less bacteria in their faeces (average Chapter 5. Black_Carbon affects on gut microbiota 138 copy number = 3199409.242; stdev = 1640755.054) (figure 5.3). The ratio of total bacteria in bees exposed to black carbon was about twice (1.8 fold change) that in control bees 5.3.

Normal PCR ampliﬁcation analysis showed ampliﬁcation and clear bands of DNA in the gel analysis for bacteria in the phlyum Actinobacteria, Firmucutes and Bacteri- odetes. But there were no DNA bands (negative gel result) of phylum: Betaprotobac- teria and Alphaprotobacteria. I performed quantitative different analysis (qPCR) between groups (control and black carbon) for those phylum showing positive DNA band in PCR analysis.

There is a reduction in bacteria belonging to Firmicutes and Bactericides in treated bees compared with untreated bees (the ratio of copy number in black carbon group to control group ; Firmcutes 0.42, Bacteriodetes 0.78) (figure 5.4). This interesting result indicates that black carbon is able to induce a change in the total and specific gut bacteria community in bumble bees, which gives potential in further studies regarding air pollution and the bee gut microbiota. There was no effect on the phylum Actinobac- teria (ratio = 1.02,) between groups. This mean probably bacterial species belong to phylum Gamma and Beta protobacteria increased. But I did not get successful PCR amplification from Gamma and Beta protobacteria in my samples. Chapter 5. Black_Carbon affects on gut microbiota 139

Absolute quantification of total bacteria 8.E+06

7.E+06

6.E+06 5863132.796

5.E+06 Copy number Copy

4.E+06

3199409.242

3.E+06

2.E+06

1.E+06

0.E+00 Control Black carbon

FIGURE 5.3: The ratio of total bacterial population in bees faeces between treatment and controls. A ratio below 1 indicates less bacterial population in control bees compared to those treated with black carbon.

FIGURE 5.4: The ratio of bacterial population in faeces between treatments (black carbon and control) of non reproductive workers Bombus terrestris. Charts ratio less than 1 indicates there were less bacteria in treated compare with untreated bees and equal to 1 mean no difference. Chapter 5. Black_Carbon affects on gut microbiota 140

5.3.4 Bioinformatic metagenomic analysis

32 libraries of 16S rRNA gene (HVR V1 - V2) were generated on an Ion Torrent platform. A total of 3649996 sequences (minimum 762, mean 114062.375, median 122712.0 and maximum 274617) with length 400 bp were obtained from 32 libraries (12 control vs 12 black carbon) and 4 library for each PCR and kit negative control. QC analysis showed that high range of quality value across most bases from 5 ﬁrst bp to 325 bp (quality score above 25 phred score). Base calls towards the end of reads are of low quality (quality score at last 100 bp less than 20).

The low quality sequences were removed leaving 166,587 sequences to construct the feature table. Feature table (OTU picking) at the 97% similarity between reads gave 280 amplicon sequence variants (features or OTU or phylotypes or species) across all samples. Samples and features that show up only one or a few reads and features existing in 2 or less samples were removed and sampling depths across different samples was adjusted to least to 1350 reads. The 24 samples returned in count table and reads have come from PCR and kit negative control were excluded for further analysis. Thereby, feature count (species) have been decreased to 96 features and these features were passed for further analysis (alpha, beta and composition diversity).

5.3.5 Taxonomy and relative abundance

The three most abundant phyla in our dataset were: Firmicutes, Actinobacteria and Gammaprotobacteria (figure 5.5). In some samples, there were rare sequences (about 1.5 %) belonging to phylum Bacteriodes (figure 5.5, red section charts). Surprisingly, high abundance of phylum Bacteriodes was observed in three samples (Control pre treated 24 %; black carbon pre treated 71.4%; black carbon post treated 83.2%). The abundance of Phlyum Bacteriodes in these three samples disagrees with my results in chapter 3 and most studies about bacterial species in bees’ guts [1]. Therefore, these three samples were excluded form calculation average of relative abundance. The most Chapter 5. Black_Carbon affects on gut microbiota 141 striking result to emerge from the data is that there were no sequences belong to phylum Alphaprotobacteria and Betaprotobacteria agreeing with the results I obtained from normal PCR and real time qPCR analysis amplification.

From the chart (Figure 5.5; grey charts) shows, on average, 62.1 % (lowest 57.8 %, highest 64.1 %) of sequences with high abundance are displayed as phylum Firmicutes. Comparing the pre and post in both control and black carbon exposure, it can be seen that there were little change (about 0.8 % ) in abundance of phylum Firmicutes in the same bees between time points (pre 63.4 % (lowest 53.2 %, highest 74.9% - post 64.1 % (lowest 61.2 % , highest 67.3 %) while an reduction (about 5.4 %) in relative abundance was detected after bees were exposed to black carbon (before exposure = 63.2 (lowest 56, highest 70.1 - after exposure 57.8 (lowest 47.8, highest 63.8). This reduction also agrees with my qPCR data, which showed that black carbon caused a sig- niﬁcant decrease in the bacterial population of Firmicutes. The most representative genera were; Lactobacillus apis, Ruminococcaceae UCG-014, Staphylococcus, Lach- nospiraceae, Ruminococcaceae, Marvinbryantia, Streptococcus, Lactobacillus bombicola, Lachnoclostridium and Lactobacillus murinus.

The second most abundant population of bacterial species in our bees belonged to the phylum Gammaprotobacteria averaging 20.4 % (lowest 19 %, highest 23.9 ﬁgure 5.5; orange section of charts). The result showed high stability in Gammaprotobacteria population (before 19.8 (lowest 14%, highest 34.1 %) - after 19.1% (lowest 14.7, highest 28.1) in both time points in control bees. In contrast, there were an increase (about 3.4 %) in Actinobacteria population in bees that treated with black carbon (pre 19.05 (lowest 15.2 %, highest 28.1 - post exposure 23.5 (lowest 21.1, highest 33.6 %). The most representative genera were detected in this phylum were: Gilliamella, Luteimonas, Es- cherichia Shigella, Snodgrassella, Stenotrophomonas, Burkholderia, Cupriavidus and Xanthomonadaceae.

On average 12.5 % (lowest 11.7 %, highest 13.6 %) of sequences were belong to phylum Chapter 5. Black_Carbon affects on gut microbiota 142

Actinobacteria. There were stability in Actinobacteria DNA sequences between control (pre 11.7 % (lowest 8.9 %, highest 15.2 %) - post 12.4 % (lowest 7.6 %, highest 17.1 %) and treated bees with black carbon (pre 12.3 % (lowest 5.2 %, highest 18.2) - post 13.6 % (lowest 10.4%, 16.8 highest %). Interestingly, there were also no differences in the ratios of Actibobateria between control and black carbon groups in qPCR analysis. The most representative family in the phylum were; Microbacteriaceae, Intraspo- rangiaceae; genus Knoellia, Blastocatellaceae; genus Stenotrophobacter, Biﬁdobacteri- aceae; genus Biﬁdobacterium and Bombiscardovia, Ilumatobacteraceae; genus Iluma- tobacter, Propionibacteriaceae; genus Cutibacterium, Micrococcaceae; genus Kocuria palustris, Propionibacteriaceae; genus Cutibacterium and Tessaracoccus, Actinomyc- etaceae; genus Varibaculum, Intrasporangiaceae; genus Ornithinicoccus, Intrasporan- giaceae; genus Ornithinibacter; Brevibacteriaceae; genus Brevibacterium, Microbacte- riaceae; genus Microbacterium and Blastocatellaceae; genus Blastocatella. Chapter 5. Black_Carbon affects on gut microbiota 143

FIGURE 5.5: Taxonomic level plot based on the 16S amplicon sequencing, calculated according to total relative abundance across the sample set. x axis is sample - y axis is relative abundance. Con_Pre control bees before treated (sugar water), Cont_Post control bees after 2 days treated with sugar water, BC-Pre bees before expsure to black carbon and BC_Post bees after treated with black carbon. The minor phylum which them abundance less 1 % are no showing in plot. grey bars phylum Firmicutes, , brown phylum Actinobacteria, blue phylum Gammaprotobacteria, and yellow assigned as unidentiﬁed bacteria Chapter 5. Black_Carbon affects on gut microbiota 144

5.3.5.1 Diversity Indices

I checked if black carbon altered the number and proportion of bacterial species in individuals. I investigated differences in alpha diversity across all samples (black carbon (pre and post) and control (pre and post) using three alpha diversity measurements. The results showed no differences in the alpha diversity in any of these three analysis (Faith’s phylogenetic diversity; Kruskal-Wallis (pairwise) test; q value = 0.58, (ﬁg- ure 5.6), measure of community evenness (Pielou’s index) differences between groups; Kruskal-Wallis (pairwise); q value 0.43, (ﬁgure 5.7) and quantitative measure of community richness using Shannon’s diversity index; Kruskal-Wallis (pairwise); q value = 0.626207. I concluded that no association between black carbon and phylotypes richness and evenness, suggesting that black carbon was not a strong driver of variation in

bumble bees’ microbial communities. Faith’s Phylogenetic Diversity Faith’s

Black carbon post exposure Black carbon pre exposure Control post exposure Control pre exposure

Condition

FIGURE 5.6: Boxplots showing Faith’s phylogenetic diversity (alpha-diversity indexes) of bacterial communities among different replicates of black carbon (pre and post) and control (pre and post) samples. X axis indicates groups; Y axis indicates proportional of similarity bacterial species between samples.

Chapter 5. Black_Carbon affects on gut microbiota 145

evenness index evenness Pielou’s

Control post exposure Control pre exposure Control post exposure Control pre exposure

Condition

FIGURE 5.7: Boxplots show measure of community evenness (alpha-diversity indexes) in total difference proportions of bacterial species present in bumble bees. X axis indicates groups; Y axis indicates proportional of similarity bacterial species between samples

A Beta diversity analysis was performed to measure the overall change in the whole community (black carbon vs control). This was performed by comparing multivariate statistics (weighted UniFrac distance, unweighted UniFrac distance and Jaccard distance). In all analysis, I did not ﬁnd a strong difference between groups in gut bacteria community between black carbon and control group (either in pre and post treatment). As shown in PCoA Plot (ﬁgure 5.8), weighted UniFrac showed that there were no clear differences of minor or low species in abundance between treatment groups. Pairwise permanova test showed that no difference in rare species between groups (q value = 0.861).

Also, I graphically represented other similarities between individuals in groups and differences between groups using PCoA based on Jaccard similarity index. This was generated based on rare species (individual features that are different but are closely related) in the data in both black carbon and control groups. Result showed clear clustering between samples in all groups (ﬁgure 5.9). Chapter 5. Black_Carbon affects on gut microbiota 146

FIGURE 5.8: 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points black carbon pre; red points black carbon post; green points control pre; yellow points control post). Distances between samples were calculated using weighted UniFrac distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees . Chapter 5. Black_Carbon affects on gut microbiota 147

FIGURE 5.9: 3D PCoA Plot showing a sample-by-sample distance, each point represents one of the samples (blue points black carbon pre; red points black carbon post; green points control pre; yellow points control post). Distances between samples were calculated using Jaccard distance matrix. Samples close to each other means that those samples have abundance species with very similar overall phylogenetic trees .

5.3.6 Differential abundance analysis

Balance in sequences counts between species was quantified from each samples (control and black carbon (pre and post)) then a multivariate response linear regression test was performed for each of the balances. Log fold changes was calculated for coefficient p values. As shown in Heatmap tree (5.10), there is no differences in the taxa abundance between groups. The taxa (96 species) divided into 10 partitions, the first top partition was 8 species in denominator (figure 5.11) divided by 88 species in numerator, balance in sequence count for the first partition between groups did not show clear differences . As shown in (figure 5.12) bacterial species (6 Snodgrassela, 1 unclassified bacteria and 1 Chapter 5. Black_Carbon affects on gut microbiota 148

Bifidobacteriaceae) in bees treated on average partially decreased compare with all the other three groups. Sequences assigned as unclassified bacteria in Silva database while blasting against NCBI database assigned as Lactobacillus. On the other hand, composition of YO denominator (88 species (13 Gilliamela, 11 Snodgrassela, 10 unclassified bacteria, 10 Bifidobacteriaceae and 9 lactobacillus ) were increase in bees that treated with black carbon compare with other three groups. Chapter 5. Black_Carbon affects on gut microbiota 149

BC_Post BC_Pre Con_Post Con_Pre y0 y1 y2 y3 y4 y5 y6 y7 y8 y9 BeeGroup

FIGURE 5.10: Heatmap display log of the coefﬁcient p-values for each of the balances. The rows of the heatmap represent samples and the columns of the heatmap represent balances. The colour scale represents microbial abundances at log ration of p value linear regression on the balance. Each of the tips corresponds to a species, Y represents splitting the data into partitions, for example Y0 is partitions between long branches (sub trees) and short branches (sub trees). light red in Yo balances is numerators for each balance and dark red is denominators. . Chapter 5. Black_Carbon affects on gut microbiota 150

y0numerator taxa (88 taxa)

D_5__Gilliamella

D_5__Snodgrassella

D_0__Bacteria

D_4__Bifidobacteriaceae

D_5__Lactobacillus

0 2 4 6 8 10 12

y0denominator taxa (8 taxa)

D_5__Snodgrassella

D_0__Bacteria

D_4__Bifidobacteriaceae

0 2 4 6 8 10 12 Number of unique taxa

FIGURE 5.11: Shows number of unique taxa (phylotype) in both numerator and denominator partitions

y0 y0 = ln numerator y0denominator

BC_Post

Con_Pre log ratio Con_Post

BC_Pre

12 10 8 6 4 2 0 2 4

FIGURE 5.12: Boxplots showing balance taxa summary, x axis represents average the log ratio for each group, at each balance, calculated the isometric log ratio transform ; y axis BC_Post: bees after treated black carbon, BC_Pre: bees before treated with black carbon, Con_Post; Bees before treated with sugar water, Con_Pre; bees after treated with sugar water Chapter 5. Black_Carbon affects on gut microbiota 151

5.4 Discussion

Black carbon is now a common containment in the environment [11] and its effects have been identiﬁed in many biological models and wildlife [360, 361]. Recently, black carbon has been shown to affect the gut microbiota of mammals [362]. Considerable levels of particulate matter has been detected on the bodies of honey bees [15]. No- tably, bees become directly exposed to airborne particulates when they are engaged in food collection. I hypothesized that bees may have altered gut microbiota caused by exposure to particulate matter. The link between black carbon as an environmental stressor, and bee health has not been established yet. Therefore, this study set out with the aim of assessing the effect of black carbon on gut microbiota in worker Bombus terrestris using simple culture, qPCR and 16s rRNA metagenetic analysis.

To date, one study has evaluated the effect of particulate matter on diversity (alpha and beta) of bacterial community by using 16s rRNA metagenetic analysis in mice [362]. In our 16s dataset analysis, we did not observe any signiﬁcant differences in alpha and beta diversity richness between groups. This result differs from Mutlu et al. which estimated a signiﬁcant effect of particulate matter in global bacterial composition throughout the mouse gastrointestinal tract. The differences between our study and this study, could be explained by several factors including the differences in (1) in composition of gut microbiota (2) genetic background (3) response to particulate matter (4) route and period of exposure (5) particulate matter and black carbon (6) life style and (7) behavioural task. This lack of effect of black carbon on bee gut microbiota could be due to social interactions (such as trophallaxis behaviour task) between individuals in social insects, such as termites, ants, and certain bees and wasps [350]. This facilitation of transmission of gut-associated bacteria are less in other model organism. Therefore, it could be this interaction effects on bacterial species (alpha and beta) diversity between individuals.

The results of culture technique and qPCR approach showed that black carbon is able to induce certain change in total abundance of bacterial composition in the worker of Chapter 5. Black_Carbon affects on gut microbiota 152

Bombus terrestris. This was confirmed with both culturing (figure 5.2) and qPCR methods (figure 5.3) as well as 16s composition analysis showing little increase (median ratio) of 88 bacterial species in faeces of bees exposed to black carbon (figure 5.12). How- ever, collaborative work between myself and Revekka Dimou (master student) did not find any effect of black carbon on gut bacteria in the Bombus terrestris adult workers. But, there are similarities between the attitudes effect of black carbon by increasing bacterial population in bees faeces in this study and significant increases gut microbial population in the small bowel, colon and the faeces of mice as effect of particulate matter [362]. In humans particulate matter can cause intestinal disease and access to gastrointestinal and digestive system via multiple mechanisms [363]. Likewise, it may happen in bees when a significant portion of particulate matter inhaled into his gut during collection foods, Negri et al. detected substantial deposition of particulate matter in the bees gut. In humans, particulate matter have a direct effect on epithelial cells and it has been shown to enhance para cellular permeability in alveolar epithelial cell monolayers [362]. Link between Gammaprotobacteria and Firmicutes species with the host epithelium where they form biofilm-like structures was detected in worker honey [350].

Metagenomic analysis showed little reduction in relative abundance (about 4.5 %) of Firmicutes phylum in bees exposed to black carbon (ﬁgure 5.5). In mice individuals exposed to heavy metals have an increase in the relative abundance of Firmicutes [364]. Fruit ﬂy larvae exposed to nano particles show a less diverse gut microbiota and over- growth of Lactobacillus brevis [365]. In contrast, the ratio of Firmicutes phylum in this study decreased as effect of black carbon. However, Firmicutes phylum in our bees gut was recorded at high abundance 62.1 % which is consistent with previous studies in bumble bee [266] and honey bee [1].

In 16s rRNA experiment, 62.1 % of sequences were assigned to lactic acid bacteria: Fir- micutes phylum with three different species of Lactobacillus which previously were identiﬁed in the bumble bee [1]. Contrary to expectations, in simple culture experiment we did not observe a growth of Lactobacillus on both BHI and MRS media. This Chapter 5. Black_Carbon affects on gut microbiota 153

ﬁnding is contrary to previous studies which have found that both BHI and MRS media provide best condition to growth bees gut lactic acid bacteria ( Lactobacillus)[366]. This rather contradictory result may be due to different in biological samples (different colonies) [1]; there were differences in bees ages between ﬁrst experiment (culture method) and second experiment. There is a remarkable difference in the gut microbiome between the different bees castes: callow worker (zero), the female workers, the forages and the male drones [367]. Other possible factors could be due to potential experimental errors.

In addition, all three methods confirm that the most abundant phyla in the faeces of bumble bee were Firmicutes, Actinobacteria, Gammaprotobacteria and Bacteriodetes. It is somewhat surprising that we did not find bacterial species belonging to the phyla Alphaprotobacteria and Betaprotobacteria in the faeces worker bumble bee. These results differ from results in chapter 3 and Kwong and Moran conclusion that Alphapro- tobacteria and Betaprotobacteria are a common phyla in the bumble bee’s gut. The absence of these 2 phyla in our bees could be due to different sampling tissue in this study (faeces) and previous studies they used gut tissue. Moreover, one unanticipated finding was that we did not observe Wolbachia DNA sequences in our bees faeces. This is an important point for future research about exploring whole shotgun meta-genomic sequencing in the bees. Wolbachia infections are estimated to be widespread among arthropods, causing big obstacles to researchers in performing whole shotgun metagenomic sequencing in bees.

Overall, the results of this research support the idea that black carbon has the potential impact on gut microbiota. The most obvious ﬁnding to emerge from this study is that black carbon lead an increase the total bacterial species whereas an reduction in phylum Firmicutes abundance in worker Bombus terrestris. The results of next generation sequencing of 16s rRNA gene showed faeces contain 96 different strains. We did not observe Wolbachia species in faeces samples and this would be a fruitful area for further work in performing whole shotgun meta-genomic sequencing in the bees. Chapter 5. Black_Carbon affects on gut microbiota 154

5.5 Collaborative work statement

In ﬁrst experiment bee husbandry, treatment bees with black carbon, cultivation of bacteria, PCR and real time qPCR were carried out by myself. Dr Caroline Cayrou in lab 121, Department of Genetics and Genome Biology, University of Leicester assisted me in the PCR approach. Dr Mirko Pegoraro lab 219, at same department assisted me in the real time qPCR and data analysis. In second experiment, bee husbandry, treatment bees with black carbon, PCR and real time qPCR were conducted together by Revekka and myself while analysis qPCR data and preparing DNA materials for next generation sequencing data were conducted by Revekka Dimou. I conducted all the bioinformatic analysis. Chapter 6

Discussing ﬁndings

6.1 A summary of the chapters result

In this thesis, my main focus was on the effects of the neonicotinoid, imidacloprid on insects. During my studies, I found imidacloprid to have numerous effects on non reproductive workers of the buff tailed bumble bee Bombus terrestris audax. My thesis tried to understand the mechanism of these effects by looking at imidacloprid effects on DNA methylation, gene expression, gut microbiota and behaviour using the Drosophila melanogaster model. As an extension of this work, I researched the effect of black carbon on gut microbiota in the bumble bee. Below is a quick summary of my main ﬁndings.

Chapter 2 I found 79, 86 and 16 genes differentially methylated at CpGs, CHHs and CHGs sites respectively between control and bees exposed to imidacloprid. In addition, I found CpG methylation much more focussed in exons region compared with methylation at CHH or CHG sites. From the differentially methylated genes, I have found 147 (CpGs), 215 (CHHs), 54 (CHGs) enriched gene ontology terms. KEGG pathway analysis showed a mitogen-activated protein kinase (MAPK) signaling cascade

155 Chapter 7. Discussing ﬁndings 156

(PATHWAY: bter04013) was over represented in differentially methylated CpGs . High CpG methylation was associated with highly expressed genes. Neonicotinoid treated bees had higher levels of non-CpG methylation.

I found 378 genes were differentially expressed: 216 genes up regulated and 162 down regulated in neonicotinoid samples compared with controls. Six cytochrome detoxiﬁ- cation genes were differentially expressed and 4 genes in Homeobox (Hox) family were downregulated in neonicotinoid treated bees. I found 209 enriched GO terms associated with differential gene expression. A large number of these GO terms were associated with energy reserve metabolism and the negative regulation of neuromuscular synaptic transmission.

In addition, I found 25 genes differentially alternatively spliced between control and neonicotinoids samples. The most represented GO terms associated with these alternative spliced splice genes were toxin transporter activity, asparaginase activity, peptide methionine (S) -S-oxide reductase activity, facilitated trehalose transporter Tret1-2 and Cytochrome activity system.

Chapter 3 The composition of the gut microbiome in bees is very important in many biochemical process. However, it is susceptible and can be inﬂuenced or disturbed during the course of physiological and behavioural changes. Stress from pathogen and insecticides have all been shown to affect gut microbiomes. An exciting ﬁeld of research is the gut-brain axis. In chapter 2, I showed that neonicotinoid insecticides affects many genes related to the nervous and immune system (see chapter 2). Could neonicotinoids have an effect on the gut microbiome as well?

A total of 3,092,487 sequences with length 2 x 250 bp were obtained from 30 libraries. Five most abundant phylum in our sequences of 16S rRNA gene were: Betaproteobac- teria (37.45 %), Gammaproteobacteria (30 %), Actinobacteria (10.6 %) and Firmicutes (18.95 %) and rest bacteria were belong to Alphaprotobacteria , and Bacteroides. The PERMANOVA test, alpha and beta diversity analysis showed no significant effect of the Chapter 7. Discussing findings 157 imidacloprid treatment on the distribution of bacteria among samples. However, in differential abundance analysis with gneiss I found a strong inhibitory effect of imidacloprid on Lactobacillus helsingborgensis. In addiction, the causal agent of American foulbrood disease which is destructive and a notifiable pest of honey bee colonies was detected for the first time in bumble bee gut. Imidacloprid did not show antibacterial toxicity toward Escherichia coli and Staphylococcus aureus.

Chapter 4 In this chapter I tried to develop a behaviour model for neonicotinoid exposure. Circadian rhythm is a central behaviour of bees and is well studied in the model system Drosophila melanogaster.

The results showed signiﬁcant differences between strains of Drosophila melanogaster in negative geotaxis climbing ability and locomotor activity in light, dark and constant darkness conditions. Low doses of imidacloprid increased both strains activity while high doses decreased activities. Canton-S strain was more sensitive to imidacloprid during geotaxis assay than M1217. In contrast, M1217 daily locomotor activities were sensitive to imidaloprid while the Canton-S strain was not.

Chapter 5 This chapter is a bit digression to other chapters. Bees become directly exposed to airborne particulates when they are engaged in food collection. Other researchers in my department are very interested in the effect of black carbon on gut microbiota in mice. As preliminary work, to see if this is a phenomenon in insects. I tested the effect of black carbon on bumble bees gut microbiota.

Bacterial cultivation methods showed significant increases of species (CFU) in bees exposed to black carbon and this was detected in the qPCR approach as well. However, in metagenomic analysis, I did not find significant differences in alpha and beta diversity richness between black carbon and control groups. The most abundance identified phylum were Firmicutes (62.1 %), Gammaprotobacteria (20.4 %), Actinobacteria (12.5 Chapter 7. Discussing findings 158

%) and Bacteriodetes. In bees exposed to black carbon there were an reduction of relative abundance in the phylum Firmicutes by (4.5 %) and 0.4 fold change in both 16s and qpCR approaches respectively. Further analysis showed that there was no wolbachia species in faeces. These ﬁndings contribute in several ways to our understanding of gut bacterial community in bumble bee and how it is inﬂuenced by air pollution.

6.2 General discussion and conclusion

The role of insect pollinators in feeding a growing population and their recent declines is a major concern. In 2013, the European Union imposed a partial restriction on neonicotinoid use which is currently being reviewed [368]. The National Farmers Union regularly requests and is granted licenses for emergency use of neonicotinoids [81]. Out- side of Europe, few countries have introduced restrictions on neonicotinoid use and they remain the most widely used insecticides in the world.

I believe that there is no doubt about the effect of neoicotinoids on non target organisms. This thesis has provided a deeper insight into epigenetic, gene expression and gut microbiota effects of imidacloprid on bumble bees. These results lay the ground- work for future research into (1) understanding the rate of bees tolerance toward neonicotinoid insecticides, (2) exploring adverse effects of neonicotinoids over the course of many generations and (3) negative effects on non neural systems such as gut microbiota.

Bees resistance to neonicotinoids

My results in chapter 2 provide empirical evidence of existing detoxification genes (Cy- tochrome subfamily of P450s) which were disrupted by non lethal doses of imidacloprid. This result is in line with previous studies of the importance of enzymes to protect bumble bees and honey bees against the toxicity of neonicotinoids [226, 235, 369]. It Chapter 7. Discussing findings 159 can therefore be assumed that defence systems of workers bumble bee responded to non lethal dose (10 ppb) of imidacloprid. In Drosophila melanogaster an over expression of detoxification genes correlates with increased survival in resistance to pesticide, while cytochrome knock-down dramatically reduce survival [225]. More recently, Man- jon et al. claimed that both honey bee and bumble bee have biochemical defence systems (Cytochrome genes) to imidacloprid and thiacloprid [226]. However this appears not be totally successful. Several lines of evidence showed that sub lethal concentrations of neonicotinoids have adverse affects on managed bee and wild bee including, efficiency of foraging and flight ability [111, 112, 370–372], negatively affected learning and memory [118] and immunosuppressive effects [123, 373, 374].

We know that non lethal dose of neonicotionoid have multiple effects on bees health. These effects could be long lasting, either because, (1) the molecular mechanisms, physiological and behaviours processes could be inherited, (2) neonicotinoid remain in soil, water and bees food for a long time [85, 86,88,90, 91], or as mixture with other pesticides [375]. There is evidence that mixtures of pesticides has more effects than single pesticides [376–378]. For example, there are synergistic interactions between the neonicotinoid and a fungicide on survival probability in a solitary bee [377].

We can only speculate whether synergistic and/ or additive effects will happen with the introduction of a new generation of pesticides. Therefore, urgent action should be taken to evaluate the potential sub-lethal effects of a new generation of pesticides which will replacing neonicotinoids. In 2014, Sulfoxaflor pesticide was licensed in many countries worldwide, including China [379], Canada [380] and Australia [381]. Europe [382]. Like neonicotinoid, it is a systemic insecticide which acts as an insect neurotoxin [383]. Surprisingly, as shown in neonicotinoids, there is evidence of residues of sulfoxaflor after agricultural application [384]. Alarmingly, sub lethal dose of sulfoxaflor has a negative effect on reproductive biology of bumble bee (Bombus terrestris) colonies [385]. Therefore, most likely Sulfoxaflor and residues of neonicotinoid potentially would have an added effect on pollinations. This is an important issue for Chapter 7. Discussing findings 160 future research. Included here will be any epigenetic effects, in particular DNA methylation.

Transgenerational inheritance effects

It is widely accepted that epigenetic modiﬁcations, in particular DNA methylation plays a crucial role in our understanding of the mechanisms environmental-exposure- associated destructive health effects and change biological outcomes [153, 386, 387]. Hymenopteran insects (ants, bees and wasps) are important emerging models for epigenetics [163, 246–248].

This thesis for the first time showed the effects of neonicotinoid on DNA methylation and gene expression. The impact of the neonicotinoids on bumble bees DNA methylation is important for both current bees healths and inheritance for next generation. There is evidence that environmental stress such as toxicants or abnormal nutrition play a crucial role in regulating epigenetic transgeneration inheritance of disease and phenotypic variation in plant, pigs, rats and humans [388]. There are examples of intergenerational effects due to pesticide exposure [152]. For example, exposing rats to the fungicide vinclozolin lead to effects on male reproductive functions into the F4 generation [249]. Most epigenetic reprogramming happen in germline (sperm and egg) [388]. A recent study found metabolic changes due to intergenerational epigenetics in mice could be linked to changes in methylation in the father’s sperm [389]. Sperm RNAs and sperm RNA modifications play crucial role in epigenetic inheritance [390]. Imidacloprid has effects on the weights of the epididymis, sperm parameters, and testosterone in rats [391]. Imidacloprid can negatively affect honeybee drone sperm quality [128, 171]. Furthermore, there is consensus among studies about the effect of neonicotinoid on reproductive system [125, 377, 392], Christen et al. showed that thiamethoxam has strong transcriptional effect on vitellogenin genes in the honey bee [393]. Together, all these evidence alert us that all these negative effects potentially could be extended to next generation. Chapter 7. Discussing findings 161

Intergenerational epigenetics is deﬁned as effects found in the offspring generation (F1) due to direct exposure to the stressor of the parental (F0) and/or the developing germ cells. Although there is epidemiological evidence for epigenetic inheritance, mechanisms have been lacking [389].

To discover if epigenetic effects due to neonicotinoids are intergenerational, a future suggestion is, treat the father with neonicotinoids. The father provides only sperm to the offspring colony, as unlike the mother he is not involved in the growth of the colony. This allows us to separate intergenerational epigenetic inheritance from non-epigenetic alterations in the parental metabolic milieu that cause developmental changes in the second generation [389].

Gut microbiome

If neonicotinoids have a negative effect on memory function and bees health, it therefore also likely that side effects are linked to somewhere else such as gut microbime. Be- cause there are bidirectional communication between brain and gut microbiota [132]. We proposed that imidiacloprid has a direct or indirect effect on gut micrbiota in bumble bee.

The current study found that the ratio of one species Lactobacillus extremely reduced in bees exposed to imidacloprid. This also found in the third instar larvae of Drosophila melanogaster [256]. However, the observed difference in abundance of this species was not significant in adult fruit fly or honey bee [55, 256]. A possible explanation for the differences between results, could be that due to establishment of characteristic microbiota through development stage. Bacterial colonization in bees gut are zero or less than 101 bacterial cell per gut while the abundance increase to 109 bacterial cell in adult workers [1]. I used callow workers, therefore the initial bacterial population in my study is close to larvae rather than to adult. When bacterial establishing started, the bees in Chapter 7. Discussing findings 162 treated group exposed to imidacloprid. Therefore, either, there were bees’ immune responses in after exposed to imidacloprid and then Lactobacillus population indirectly disrupted [123, 373, 374], or Lactobacillus involved in imidacloprid metabolism, absorption and biotransformation [253].

There are still many unanswered questions about the effect of neonicotinoids on bees gut micrbiota. A further study with more focus on effect through out the development stages (larvae, pupa and adult) is therefore a ﬁrst suggestion. Second, examining the toxicity of imidacloprid against Lactobacillus. Third, the role of Lactobacillus in degra- dation of imidacloprid.

Other more signiﬁcant ﬁndings was emerged from metagenomics analysis (chapter 3 and 5), is the presence or absence Wolbachia in my dataset. Different species were detected in gut tissue while faeces samples were not contained. One of the main problems of whole genome shotgun sequencing of microbial communities in insects is the preponderance of Wolbachia. Chapter 7

Future direction

Several questions, ideas, tests and experiments have been left for the future due to lack of time, for example bioinformatic analyses is can be time consuming, sometimes re- quiring days to ﬁnish a single run. Also, there are many “hidden messages” and further can be conducted on my transcriptomic and methylomic and metagenetic data. The following suggestion and ideas could be tested:

• It would be interesting to consider that neoncotinoids have inter-generational or trans-generational genetic effects on bumble bee. This could be investigated by exposing parental (F0) to imidacloprid then looking for potential effects on epigenetic processes such as DNA methylation and histon modiﬁcation. Also worthy is to look at the epigenetic effects in brain and sperm.

• A fascinating future research project is to implement candidate gene assays to look at methylation in bumble bees in the ﬁeld. Different queen will be collected from different locations. We can then carry out methyaltion analysis of candidate genes (chosen from a list of genes diferentially methylated in this study) in colonies transplanted into high and low exposure to neonicotinoid areas either in the UK or in Europe.

163 Chapter 7. Future direction 164

• More in depth analysis on gene expression data in this study (chapter2) and Colgan et al. 2019 study is useful about the pattern of expression between treated and untreated bees with neonicotinoid. In recent years, differential co- expression analyses have been increasingly used to analyse large data sets [394]. Therefore, it is worth while to perform a co-expression network analysis to show which genes have coordinated expression pattern across a group of samples (in both control and treated groups separately). We could also further investigate the alternative splicing results, for example: how do the different splice isoforms identiﬁed differ? Do they change encode different proteins? Do the differences change the functional domains in a protein? Or are functional domains missing from splice variant?

• There are a number of gaps in our knowledge around the effect of neonicotinoid on non neural parts, for example gut microbiota. The results in chapter3 investigated that imidacloprid has effect on gut microbiota and there was novel pathogeneic bacteria in guts of bumble bee. A further study will be interesting with more focus on effect through out the development stages (larvae, pupa and adult). Second, examining the toxicity of imidacloprid against Lactobacillus. Fur- ther depth investigations are required about the presence of Paenibacillus larvae in pollen which are imported to the UK.

• According to BioProject - NCBI database [395], there are few bumble bee 16s RNA microbiota data and surprisingly, there is no whole genome micrbiobio data for bumble bee or honey bee [395]. In my study I concluded that there is no Wol- bachia contamination in the faeces of bumble bee (see chapter3 and5). There- fore, I strongly encourage the performance of whole metagenome shotgun sequencing for bumble bee and honey bee from the faeces. Appendix A

Supplementary information

165 FIGURE A.1: The captured image shows the code function of pipeline to remove outlier (artefact) genes between samples in same condition. The code was generated and provided by Sascha Otts group (they ran the RNASeq on the MIBTP).

166 TABLE A.1: These tables show loc number (gene ID), transcripts and names for genes that deferentially methylated Cs (p value less than 0.01) in multiple contests (CpG, CHH, CHG) between treated and untreated bees. Diff Meth; less than 5% deferentially methylated in CHG, CHH sites and 10% in CpG sites.

Genomic Loc Diff Meth Transcript Names context

CHG LOC100644912 -11.7647058824 RNA-binding protein Musashi homolog Rbp6, transcript variant Rbp6

CHG LOC110120204 -11.1111111111 uncharacterized LOC110120204, transcript variant X1 uncharacterized protein

CHG LOC100651814 -9.756097561 LIM/homeobox protein Awh LIM/homeobox protein

CHG LOC100648152 -9.5238095238 brain protein I3 brain protein I3 167 CHG LOC100643597 -8.9285714286 synembryn-A synembryn-A

CHG LOC100650351 -7.2727272727 ras-related protein Rab-3, transcript variant X3 Rab3

CHG LOC100645391 -7.0175438596 midasin midasin

CHG LOC100644701 -6.7796610169 headcase protein, transcript variant X1 headcase protein

CHG LOC100646399 -6.6666666667 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CHG LOC100643185 -6.5573770492 diuretic hormone 44, transcript variant X3 diuretic hormone 44

CHG LOC105666194 -6.1538461538 uncharacterized LOC105666194, transcript variant X2 uncharacterized protein

CHG LOC100649775 5.4945054945 protein numb, transcript variant X3 protein numb Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHG LOC110120243 6.5573770492 LOW QUALITY PROTEIN uncharacterized protein

CHG LOC100651818 8.1632653061 RING ﬁnger protein 44, transcript variant X6 RING ﬁnger protein 44

CHG LOC100651897 10.4166666667 kazrin, transcript variant X13 kazrin

CHG LOC100649677 10.6060606061 uncharacterized LOC100649677, transcript variant X1 uncharacterized protein

168 CHH LOC100649663 -12.7659574468 serum response factor homolog, transcript variant X2 bs

CHH LOC105667086 -12.2448979592 uncharacterized protein LOC105667086 uncharacterized protein

CHH LOC110120210 -11.7647058824 uncharacterized LOC110120210 uncharacterized protein

CHH LOC100650751 -11.3636363636 uncharacterized LOC100650751, transcript variant X2 uncharacterized protein

CHH LOC100650751 -11.1111111111 uncharacterized LOC100650751, transcript variant X3 uncharacterized protein

CHH LOC100644228 -10.8108108108 hexokinase type 2, transcript variant X4 Hk2

CHH LOC100647944 -10 dipeptidase 1 DPEP1

CHH LOC100650022 -10 LOW QUALITY PROTEIN: ATP-dependent RNA helicase kurz kz Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC100645685 -9.756097561 lachesin, transcript variant X2 lachesin

CHH LOC100649308 -9.5238095238 roundabout homolog 2 Robo2

CHH LOC100651472 -9.5238095238 LOW QUALITY PROTEIN: diacylglycerol kinase eta DGKH

CHH LOC110119762 -9.5238095238 uncharacterized LOC110119762 uncharacterized protein

169 CHH LOC100647443 -9.3023255814 zinc ﬁnger protein 800, transcript variant X2 ZNF800

CHH LOC100642892 -9.3023255814 trifunctional enzyme subunit alpha, mitochondrial isoform X1 HADHA

CHH LOC100643185 -9.2307692308 diuretic hormone 44, transcript variant X1 Dh44

CHH LOC100647319 -9.0909090909 neuropeptide SIFamide receptor SIFaR

CHH LOC100648262 -8.8888888889 uncharacterized LOC100648262, transcript variant X2 uncharacterized protein

CHH LOC100646729 -8.8888888889 fatty acyl-CoA reductase 1-like FAR1

CHH LOC100650469 -8.5106382979 serine/threonine-protein kinase mig-15, transcript variant X12 mig-15

CHH LOC110119178 -8.4745762712 uncharacterized LOC110119178, transcript variant X1 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC100642510 -8.3333333333 E3 ubiquitin-protein ligase Su(dx) Su(dx)

CHH LOC100648695 -8.1967213115 low-density lipoprotein receptor-related protein 6 LRP6

CHH LOC105666500 -8.1632653061 uncharacterized LOC105666500 uncharacterized protein

CHH LOC100646726 -8.1632653061 protein neuralized, transcript variant X6 neur

170 solute carrier family 2, facilitated glucose transporter member 1, CHH LOC100642581 -8 SLC2A1 transcript variant X1

CHH LOC100646518 -8 uncharacterized LOC100646518 uncharacterized protein

CHH LOC100649869 -8 lachesin lachesin

CHH LOC100642498 -7.8947368421 teneurin-m, transcript variant X1 ten-m

Low quality protein: adhesion G protein-coupled CHH LOC100650097 -7.8431372549 ADGRA3 receptor A3

CHH LOC110119920 -7.6923076923 uncharacterized LOC110119920 uncharacterized protein

CHH LOC110120238 -7.5757575758 uncharacterized LOC110120238, transcript variant X1 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC100646561 -7.4074074074 zinc ﬁnger protein 521, transcript variant X3 ZNF521

CHH LOC100644912 -7.4074074074 RNA-binding protein Musashi homolog Rbp6, transcript variant Rbp6

CHH LOC100646518 -7.3529411765 uncharacterized LOC100646518 uncharacterized protein

CHH LOC100642498 -7.3170731707 teneurin-m, transcript variant X1 ten-m

171 signal transducer and activator of transcription C-like, transcript CHH LOC105666599 -7.2463768116 STAT C-like variant X3

CHH LOC100651021 -7.1428571429 protein TANC2, transcript variant X5 TANC2

CHH LOC100650702 -7.1428571429 dopamine receptor 2, transcript variant X2 Drd2

CHH LOC100643586 -7.0175438596 tyrosine-protein phosphatase 10D, transcript variant X2 Ptp10D

CHH LOC100648987 -6.8965517241 acetylcholine receptor subunit alpha-like, transcript variant X1 CHRNA1

CHH LOC100644824 -6.8965517241 max dimerization protein 1, transcript variant X11 MXD1

CHH LOC110120048 -6.8965517241 uncharacterized LOC110120048 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC100646726 -6.6666666667 protein neuralized, transcript variant X6 neur

CHH LOC100650668 -6.5573770492 dentin sialophosphoprotein, transcript variant X8 DSPP

CHH LOC100643464 -6.4516129032 ubiquitin carboxyl-terminal hydrolase 48 Usp48

CHH LOC100650177 -6.4516129032 uncharacterized LOC100650177 uncharacterized protein

172 CHH LOC100648612 -6.3492063492 uncharacterized protein LOC100648612 uncharacterized protein

CHH LOC100644811 -6.3291139241 neuroligin-4, Y-linked NLGN4Y

CHH LOC100643543 -5.9701492537 solute carrier family 28 member 3 isoform X2 SLC28A3

CHH LOC100646399 -5.7971014493 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CHH LOC100648957 -5.7142857143 D-3-phosphoglycerate dehydrogenase PHGDH

CHH LOC100646322 -5.4054054054 uncharacterized LOC100646322, transcript variant X6 uncharacterized protein

CHH LOC105667085 -5.1020408163 uncharacterized protein LOC105667085 uncharacterized protein

CHH LOC105667086 5.0632911392 uncharacterized protein LOC105667086 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC105667086 5.3333333333 uncharacterized protein LOC105667086 uncharacterized protein

CHH LOC105667086 5.4794520548 uncharacterized protein LOC105667086 uncharacterized protein

phosphatidylinositol-binding clathrin assembly protein LAP, CHH LOC100645783 5.4945054945 PICALM transcript variant X11

CHH LOC105666996 5.4945054945 uncharacterized LOC105666996 uncharacterized protein 173

CHH LOC105667080 5.6179775281 uncharacterized LOC105667080, transcript variant X5 uncharacterized protein

CHH LOC105666711 6.25 tyrosine-protein kinase Btk29A, transcript variant X1 Btk29A

CHH LOC100643830 6.4935064935 angiotensin-converting enzyme, transcript variant X3 ACE

CHH LOC105667086 6.9444444444 uncharacterized protein LOC105667086 uncharacterized protein

CHH LOC100651177 7.4626865672 serine proteinase stubble, transcript variant X1 Sb

CHH LOC100649903 7.9365079365 WD repeat-containing protein 43 WDR43

CHH LOC100645811 8.064516129 DDB1- and CUL4-associated factor 10 isoform X2 DCAF10 Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC100647442 8.1632653061 mitochondrial 2-oxodicarboxylate carrier, transcript variant X3 SLC25A21

CHH LOC100646399 8.1632653061 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CHH LOC100649869 8.3333333333 lachesin lachesin

CHH LOC100649124 8.3333333333 sex-lethal homolog Sxl

174 CHH LOC100643365 8.3333333333 macrophage mannose receptor 1, transcript variant X2 MRC1

CHH LOC110120238 8.5106382979 uncharacterized LOC110120238, transcript variant X1 uncharacterized protein

CHH LOC100642498 8.5106382979 teneurin-m, transcript variant X1 ten-m

CHH LOC100647914 8.5106382979 serine/threonine-protein kinase Doa, transcript variant X2 Doa

CHH LOC100650469 8.6956521739 serine/threonine-protein kinase mig-15, transcript variant X12 mig-15

CHH LOC100649948 9.0909090909 uncharacterized LOC100649948, transcript variant X5 uncharacterized protein

CHH LOC105666194 9.2592592593 uncharacterized LOC105666194, transcript variant X2 uncharacterized protein

CHH LOC100646729 9.4339622642 fatty acyl-CoA reductase 1-like FAR1 Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CHH LOC105667085 9.5238095238 uncharacterized protein LOC105667085 uncharacterized protein

CHH LOC105666599 9.756097561 signal transducer and activator of transcription C-like STAT C-like

CHH LOC105666600 10 short transient receptor potential channel 4-like TRPC4

CHH LOC105667085 10.2564102564 uncharacterized protein LOC105667085 uncharacterized protein

175 CHH LOC100650057 10.6382978723 colorectal mutant cancer protein, transcript variant X2 MCC

CHH LOC100643626 10.6382978723 luc7-like protein 3 LUC7L3

CHH LOC100645803 10.8695652174 nuclear factor 1 X-type, transcript variant X6 NFIX

CHH LOC100647823 10.9090909091 poly(rC)-binding protein 3, transcript variant X2 PCBP3

CHH LOC100644701 11.1111111111 headcase protein, transcript variant X2 hdc

CpG LOC100645350 -38.9915966387 uncharacterized protein LOC100645350 uncharacterized protein

CpG LOC100649677 -34.5514950166 uncharacterized protein LOC100649677 isoform X1 uncharacterized protein

CpG LOC100643219 -29.7779076563 putative pre-mRNA-splicing factor ATP-dependent RNA helicase PRP1 Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100646606 -28.3088235294 ras-related and estrogen-regulated growth inhibitor RERG

CpG LOC100651177 -27.9434850863 serine proteinase stubble, transcript X3 Sb

CpG LOC100649677 -26.6666666667 uncharacterized LOC100649677, transcript variant X1 uncharacterized protein

CpG LOC100647000 -26.6441005803 dedicator of cytokinesis protein 1 Dock1

176 CpG LOC100646122 -26.4444444444 P protein P protein

CpG LOC100651463 -26.0712566201 uncharacterized LOC100651463, transcript variant X1 uncharacterized protein

CpG LOC100645140 -25.9590792839 protein phosphatase PP2A 55 kDa regulatory subunit tws

CpG LOC100643350 -25.8712121212 myotubularin-related protein 6 isoform X1 MTMR6

CpG LOC100645971 -25.4154447703 low-density lipoprotein receptor-related protein 2 LRP2

CpG LOC100647000 -24.5036868973 dedicator of cytokinesis protein 1 Dock1

CpG LOC105667085 -24.3580337491 uncharacterized protein LOC105667085 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

ras-related and estrogen-regulated growth inhibitor, transcript CpG LOC100646606 -24.1586538462 RERG variant X2

CpG LOC100643555 -23.7296747967 UBA-like domain-containing protein 2, transcript variant X1 UBALD2

CpG LOC100650059 -22.5197541703 uncharacterized protein LOC100650059 uncharacterized protein

CpG LOC100643555 -22.261663286 UBA-like domain-containing protein 2, transcript variant X2 UBALD2 177

CpG LOC100646399 -21.7345872518 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100650469 -20.8457415128 serine/threonine-protein kinase mig-15, transcript variant X12 mig-15

CpG LOC100648918 -20.7207207207 LOW QUALITY PROTEIN: mucin-3A mucin-3A

CpG LOC100644466 -20.5882352941 uncharacterized LOC100644466, transcript variant X2 uncharacterized protein

CpG LOC100646399 -20.5056179775 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100643228 -20.2959830867 B-box type zinc ﬁnger protein ncl-1 ncl-1

CpG LOC100646011 -19.4444444444 dynein heavy chain 2, axonemal Dnah2 Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100646399 -19.0564292322 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100646089 -18.5283018868 uncharacterized LOC100646089 uncharacterized protein

CpG LOC100643350 -18.4210526316 myotubularin-related protein 6, transcript variant X1 MTMR6

CpG LOC100644148 -17.7000529942 allatostatin-A receptor, transcript variant X10 AstaA-R

178 CpG LOC100646366 -16.71785871 protein decapentaplegic, transcript variant X1 dpp

CpG LOC100646399 -16.5760869565 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100646042 -16.2962962963 protein krasavietz kra

CpG LOC100648846 -16.2264150943 L-selectin L-selectin

CpG LOC100646399 -15.908045977 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100647879 -15.9019484601 cyclic nucleotide-gated cation channel alpha-3, transcript variant CNGA3

CpG LOC110120048 -15.3846153846 uncharacterized LOC110120048 uncharacterized protein

CpG LOC100650059 -15.2852049911 uncharacterized protein LOC100650059 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100645140 -15.1515151515 protein phosphatase PP2A 55 kDa regulatory, transcript variant tws

CpG LOC100651145 -14.6341463415 ets DNA-binding protein pokkuri, transcript variant X1 aop

CpG LOC100644824 -14.5833333333 max dimerization protein 1, transcript variant X11 MXD1

CpG LOC100647885 -14.3228200371 E3 ubiquitin-protein ligase MIB2 MIB2

179 CpG LOC100648308 -14.2857142857 neurogenic locus notch homolog protein 1 Notch1

CpG LOC100646399 -13.8888888889 uncharacterized LOC100646399, transcript variant X2 uncharacterized protein

CpG LOC100648496 -13.6363636364 repressed by EFG1 protein 1, transcript variant X5 RBE1

CpG LOC100643099 -13.3333333333 signal recognition particle subunit SRP72 SRP72

CpG LOC100642229 -13.1362889984 macoilin-1, transcript variant X5 macoilin-1

CpG LOC110120039 -12.5 uncharacterized LOC110120039, transcript variant X2 uncharacterized protein

CpG LOC110120039 -12.5 uncharacterized LOC110120039, transcript variant X4 uncharacterized protein

CpG LOC100646486 -12.1951219512 cilia- and ﬂagella-associated protein 61 CFAP61 Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100650105 -12.1951219512 Golgi integral membrane protein 4 isoform X1 GOLIM4

CpG LOC100643342 -11.7647058824 zinc ﬁnger protein rotund, transcript variant X1 rn

putative polypeptide N-acetylgalactosaminyltransferase 9, CpG LOC100649855 -11.4285714286 pgant9 transcript variant X4

CpG LOC100649677 -11.320754717 uncharacterized LOC100649677, transcript variant X1 uncharacterized protein 180

CpG LOC100651897 -11.1842105263 kazrin, transcript variant X13 KAZN

CpG LOC110119769 -11.1111111111 uncharacterized LOC110119769 uncharacterized protein

CpG LOC100647316 -10.9090909091 complexin, transcript variant X1 complexin

CpG LOC100649308 -10.5263157895 roundabout homolog 2 ROBO2

sodium-coupled monocarboxylate transporter 1, transcript CpG LOC105666567 -10.3448275862 SLC5A8 variant X2

CpG LOC100642498 10.2564102564 teneurin-m, transcript variant X1 ten-m

CpG LOC100651590 10.5743424584 cadherin-87A, transcript variant X1 cadherin-87A Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100642625 10.6382978723 PDF receptor, transcript variant X7 Pdfr

CpG LOC100647201 11.2903225806 alpha-2B adrenergic receptor ADRA2B

CpG LOC100643224 11.5384615385 homeotic protein Sex combs reduced Scr

brefeldin A-inhibited guanine nucleotide-exchange protein 1, CpG LOC100648040 11.7647058824 ARFGEF1 transcript variant X1 181

CpG LOC100649421 11.8476190476 uncharacterized LOC100649421, transcript variant X2 uncharacterized protein

CpG LOC100651897 11.941724385 kazrin, transcript variant X13 KAZN

CpG LOC100644824 12.2655122655 max dimerization protein 1, transcript variant X11 MXD1

CpG LOC100643467 12.7659574468 LOW QUALITY PROTEIN uncharacterized protein

CpG LOC100645391 13.3333333333 midasin midasin

CpG LOC100649604 13.9534883721 uncharacterized protein LOC100649604 isoform X1 uncharacterized protein

CpG LOC100649604 13.9534883721 uncharacterized protein LOC100649604 isoform X1 uncharacterized protein Table A.1 continued from previous page

Genomic Loc DiffMeth Transcript Names context

CpG LOC100651145 14.2857142857 ets DNA-binding protein pokkuri, transcript variant X1 aop

CpG LOC100642392 15.7894736842 uncharacterized LOC100642392 uncharacterized protein

CpG LOC100645296 18.1818181818 phosphoinositide 3-kinase adapter protein 1 PIK3AP1

CpG LOC100642540 19.391025641 attractin-like protein 1 ATRNL1

182 CpG LOC100649869 22.5870646766 lachesin lachesin

CpG LOC100644028 24.2864556305 endophilin-A, transcript variant X2 EndoA

CpG LOC100651177 25.175315568 serine proteinase stubble, transcript variant X1 Sb

CpG LOC100645476 26.9565217391 uncharacterized protein LOC100645476 isoform X2 uncharacterized protein TABLE A.2: These tables show gene ontology terms and biological process for genes that differentially methylated Cs in multiple contests (CpG, CHH, CHG) between treated and treated bees.

Context Gene ontology P-value Biological processes FDR

CPG GO:0048103 0.0002227 somatic stem cell division 0.026844361

CPG GO:0045597 0.000365229 positive regulation of cell differentiation 0.026844361

CHH GO:0030717 0.000972326 karyosome formation 0.030103909

CHG GO:0031630 0.001004016 regulation of synaptic vesicle fusion to presynaptic active zone membrane 0.036900464

CHH GO:0048707 0.001009843 instar larval or pupal morphogenesis 0.030103909 183

CPG GO:0007406 0.001144 negative regulation of neuroblast proliferation 0.03454073

CHH GO:0007552 0.001309806 metamorphosis 0.030103909

CPG GO:0007432 0.001424748 salivary gland boundary speciﬁcation 0.03454073

CHH GO:0007430 0.001550689 terminal branching, open tracheal system 0.030103909

CHH GO:0035209 0.001888013 pupal development 0.030103909

CHG GO:0009786 0.00200715 regulation of asymmetric cell division 0.036900464

CPG GO:0006928 0.002888814 movement of cell or subcellular component 0.03454073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0042694 0.003009402 muscle cell fate speciﬁcation 0.036900464

CPG GO:0010721 0.00303781 negative regulation of cell development 0.03454073

CPG GO:0007156 0.00308957 homophilic cell adhesion via plasma membrane adhesion molecules 0.03454073

CPG GO:0055057 0.003263877 neuroblast division 0.03454073

184 CHH GO:0007270 0.00348878 neuron-neuron synaptic transmission 0.030103909

CHH GO:0035277 0.003549432 spiracle morphogenesis, open tracheal system 0.030103909

CHH GO:0098916 0.00358009 anterograde trans-synaptic signaling 0.030103909

CHH GO:0099536 0.003651539 synaptic signaling 0.030103909

CPG GO:0060429 0.003823261 epithelium development 0.03454073

CPG GO:0051124 0.003830089 synaptic growth at neuromuscular junction 0.03454073

CPG GO:0048667 0.003916397 cell morphogenesis involved in neuron differentiation 0.03454073

CHG GO:0061320 0.004010774 pericardial nephrocyte differentiation 0.036900464 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0021537 0.004103271 telencephalon development 0.030103909

CHH GO:0007435 0.004103271 salivary gland morphogenesis 0.030103909

CPG GO:0061564 0.004695069 axon development 0.03454073

CHH GO:0040013 0.004779761 negative regulation of locomotion 0.030103909

185 CHH GO:0048667 0.004992874 cell morphogenesis involved in neuron differentiation 0.030103909

CHG GO:0048789 0.005011265 cytoskeletal matrix organization at active zone 0.036900464

CHH GO:0042749 0.005112852 regulation of circadian sleep/wake cycle 0.030103909

CHH GO:0007424 0.005542109 open tracheal system development 0.030103909

CHH GO:0032502 0.005637049 developmental process 0.030103909

CHH GO:0022410 0.005692504 circadian sleep/wake cycle process 0.030103909

CHH GO:0050772 0.005692504 positive regulation of axonogenesis 0.030103909

CPG GO:0060086 0.005773092 circadian temperature homeostasis 0.03454073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0045201 0.005773092 maintenance of neuroblast polarity 0.03454073

CPG GO:0033563 0.005773092 dorsal/ventral axon guidance 0.03454073

CPG GO:0050925 0.005773092 negative regulation of negative chemotaxis 0.03454073

CPG GO:0034154 0.005773092 toll-like receptor 7 signaling pathway 0.03454073

186 CPG GO:0034142 0.005773092 toll-like receptor 4 signaling pathway 0.03454073

CPG GO:0034134 0.005773092 toll-like receptor 2 signaling pathway 0.03454073

CPG GO:0034628 0.005773092 ’de novo’ NAD biosynthetic process from aspartate 0.03454073

CPG GO:0090284 0.005773092 positive regulation of protein glycosylation in Golgi 0.03454073

CHH GO:0061564 0.005808227 axon development 0.030103909

CHH GO:1990709 0.005903039 presynaptic active zone organization 0.030103909

CHH GO:0035154 0.005903781 terminal cell fate speciﬁcation, open tracheal system 0.030103909

CPG GO:0016337 0.005989585 single organismal cell-cell adhesion 0.03454073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0061382 0.006010877 Malpighian tubule tip cell differentiation 0.036900464

CHG GO:0048790 0.006010877 maintenance of presynaptic active zone structure 0.036900464

CHG GO:0045035 0.006010877 sensory organ precursor cell division 0.036900464

CHH GO:2000255 0.006024096 negative regulation of male germ cell proliferation 0.030103909

187 CHH GO:0001306 0.006024096 age-dependent response to oxidative stress 0.030103909

CHH GO:0005989 0.006024096 lactose biosynthetic process 0.030103909

CHH GO:1902474 0.006024096 positive regulation of protein localization to synapse 0.030103909

CHH GO:0097115 0.006024096 neurexin clustering involved in presynaptic membrane assembly 0.030103909

CHH GO:0044340 0.006024096 canonical Wnt signaling pathway involved in regulation of cell proliferation 0.030103909

CHH GO:0044335 0.006024096 canonical Wnt signaling pathway involved in neural crest cell differentiation 0.030103909

CHH GO:0015689 0.006024096 molybdate ion transport 0.030103909

CHH GO:0033563 0.006024096 dorsal/ventral axon guidance 0.030103909 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0050925 0.006024096 negative regulation of negative chemotaxis 0.030103909

CHH GO:0098942 0.006024096 retrograde trans-synaptic signaling by trans-synaptic protein complex 0.030103909

CHH GO:0034392 0.006024096 negative regulation of smooth muscle cell apoptotic process 0.030103909

CHH GO:0023041 0.006024096 neuronal signal transduction 0.030103909

188 CHH GO:0034125 0.006024096 negative regulation of MyD88-dependent toll-like receptor signaling pathway 0.030103909

CHH GO:0071397 0.006024096 cellular response to cholesterol 0.030103909

CHH GO:0042321 0.006024096 negative regulation of circadian sleep/wake cycle, sleep 0.030103909

CHH GO:1905520 0.006024096 positive regulation of presynaptic active zone assembly 0.030103909

CHH GO:1905606 0.006024096 regulation of presynapse assembly 0.030103909

CHH GO:1900029 0.006024096 positive regulation of rufﬂe assembly 0.030103909

CHH GO:1905936 0.006024096 regulation of germ cell proliferation 0.030103909

CHH GO:1904953 0.006024096 Wnt signaling pathway involved in midbrain dopaminergic neuron diff 0.030103909 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

positive regulation of Wnt signaling pathway involved in dorsal/ventral CHH GO:2000055 0.006024096 0.030103909 axis speciﬁcation

CPG GO:0048812 0.006267513 neuron projection morphogenesis 0.03454073

CHH GO:0072553 0.006300818 terminal button organization 0.030103909

CHH GO:0010171 0.006300818 body morphogenesis 0.030103909 189

CPG GO:0032502 0.006439007 developmental process 0.03454073

CPG GO:0090132 0.006501126 epithelium migration 0.03454073

CPG GO:0022610 0.006714263 biological adhesion 0.03454073

CHH GO:0007472 0.006779961 wing disc morphogenesis 0.031073963

CHH GO:0060562 0.006891114 epithelial tube morphogenesis 0.031073963

CHH GO:0035071 0.006937443 salivary gland cell autophagic cell death 0.031073963

CHG GO:0035155 0.00700961 negative regulation of terminal cell fate speciﬁcation, open tracheal system 0.036900464 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0007611 0.007163283 learning or memory 0.031430733

CPG GO:0045785 0.007231081 positive regulation of cell adhesion 0.03454073

CPG GO:0048699 0.007340095 generation of neurons 0.03454073

CHG GO:0032482 0.008007465 Rab protein signal transduction 0.036900464

190 CHG GO:0099500 0.008007465 vesicle fusion to plasma membrane 0.036900464

CPG GO:0042462 0.008132763 eye photoreceptor cell development 0.03454073

CHH GO:0021987 0.008294232 cerebral cortex development 0.034901632

CPG GO:0008038 0.008447404 neuron recognition 0.03454073

CPG GO:2000177 0.008989772 regulation of neural precursor cell proliferation 0.03454073

CHG GO:0035153 0.009004443 epithelial cell type speciﬁcation, open tracheal system 0.036900464

CHG GO:1904799 0.009004443 regulation of neuron remodeling 0.036900464

CHH GO:0035120 0.009295947 post-embryonic appendage morphogenesis 0.034901632 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0006935 0.009513185 chemotaxis 0.03454073

CPG GO:0048865 0.009703367 stem cell fate commitment 0.03454073

CPG GO:0098722 0.009703367 asymmetric stem cell division 0.03454073

CPG GO:0010634 0.009703367 positive regulation of epithelial cell migration 0.03454073

191 CHH GO:0001964 0.009760121 startle response 0.034901632

CHH GO:0008586 0.009760121 imaginal disc-derived wing vein morphogenesis 0.034901632

CHH GO:0008587 0.009760121 imaginal disc-derived wing margin morphogenesis 0.034901632

CHH GO:0048699 0.009983362 generation of neurons 0.034901632

CHG GO:0007430 0.010000543 terminal branching, open tracheal system 0.036900464

CHH GO:0048729 0.010074817 tissue morphogenesis 0.034901632

CHH GO:0035272 0.010228422 exocrine system development 0.034901632

CHH GO:0035114 0.010755737 imaginal disc-derived appendage morphogenesis 0.034901632 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0016199 0.011203558 axon midline choice point recognition 0.03454073

CHH GO:0060033 0.011332402 anatomical structure regression 0.034901632

CPG GO:2000400 0.011513577 positive regulation of thymocyte aggregation 0.03454073

CPG GO:0060385 0.011513577 axonogenesis involved in innervation 0.03454073

192 CPG GO:0060233 0.011513577 oenocyte delamination 0.03454073

CPG GO:0055060 0.011513577 asymmetric neuroblast division resulting in ganglion mother cell formation 0.03454073

CPG GO:0021891 0.011513577 olfactory bulb interneuron development 0.03454073

CPG GO:0045470 0.011513577 R8 cell-mediated photoreceptor organization 0.03454073

CPG GO:0033092 0.011513577 positive regulation of immature T cell proliferation in thymus 0.03454073

CPG GO:0033083 0.011513577 regulation of immature T cell proliferation 0.03454073

CPG GO:0045678 0.011513577 positive regulation of R7 cell differentiation 0.03454073

CPG GO:0007381 0.011513577 speciﬁcation of segmental identity, labial segment 0.03454073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0034162 0.011513577 toll-like receptor 9 signaling pathway 0.03454073

CPG GO:1990966 0.011513577 ATP generation from poly-ADP-D-ribose 0.03454073

CPG GO:2000011 0.011513577 regulation of adaxial/abaxial pattern formation 0.03454073

CPG GO:0000902 0.011749412 cell morphogenesis 0.034543272

193 CPG GO:0048149 0.011989572 behavioral response to ethanol 0.034558179

CHG GO:0007400 0.011990117 neuroblast fate determination 0.036900464

CHH GO:2000809 0.012012655 positive regulation of synaptic vesicle clustering 0.034901632

CHH GO:0060319 0.012012655 primitive erythrocyte differentiation 0.034901632

CHH GO:0021891 0.012012655 olfactory bulb interneuron development 0.034901632

CHH GO:0097113 0.012012655 AMPA glutamate receptor clustering 0.034901632

CHH GO:0097114 0.012012655 NMDA glutamate receptor clustering 0.034901632

CHH GO:0060535 0.012012655 trachea cartilage morphogenesis 0.034901632 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:1903223 0.012012655 positive regulation of oxidative stress-induced neuron death 0.034901632

CHH GO:0010841 0.012012655 positive regulation of circadian sleep/wake cycle, wakefulness 0.034901632

CHH GO:0090009 0.012012655 primitive streak formation 0.034901632

CHH GO:0046960 0.012012655 sensitization 0.034901632

194 CHH GO:0090118 0.012012655 receptor-mediated endocytosis involved in cholesterol transport 0.034901632

CHH GO:0090244 0.012012655 Wnt signaling pathway involved in somitogenesis 0.034901632

CHH GO:0090245 0.012012655 axis elongation involved in somitogenesis 0.034901632

CHH GO:1901631 0.012012655 positive regulation of presynaptic membrane organization 0.034901632

CHH GO:2000310 0.012012655 regulation of NMDA receptor activity 0.034901632

CHH GO:0042461 0.012187624 photoreceptor cell development 0.034937854

CHH GO:0006874 0.013027612 cellular calcium ion homeostasis 0.036854428

CPG GO:0045467 0.013631994 R7 cell development 0.037776582 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0098916 0.013808884 anterograde trans-synaptic signaling 0.037776582

CHH GO:0048854 0.013884512 brain morphogenesis 0.038271411

CHH GO:0090090 0.013884512 negative regulation of canonical Wnt signaling pathway 0.038271411

CHG GO:1902803 0.01397619 regulation of synaptic vesicle transport 0.036900464

195 CPG GO:0099536 0.014030773 synaptic signaling 0.037776582

CPG GO:0001667 0.014134095 ameboidal-type cell migration 0.037776582

CPG GO:0007420 0.014710948 brain development 0.038616238

CHG GO:0055059 0.014967917 asymmetric neuroblast division 0.036900464

CHG GO:0036445 0.014967917 neuronal stem cell division 0.036900464

CHH GO:0050768 0.015270069 negative regulation of neurogenesis 0.040659623

CPG GO:0072091 0.015366381 regulation of stem cell proliferation 0.038720046

CHG GO:0061318 0.01595877 renal ﬁltration cell differentiation 0.036900464 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0048172 0.01595877 regulation of short-term neuronal synaptic plasticity 0.036900464

CHG GO:0035147 0.01595877 branch fusion, open tracheal system 0.036900464

CHH GO:0031987 0.016661261 locomotion involved in locomotory behavior 0.040659623

CPG GO:0002065 0.016912305 columnar/cuboidal epithelial cell differentiation 0.038720046

196 CHG GO:0016360 0.016948751 sensory organ precursor cell fate determination 0.036900464

CHG GO:0016246 0.016948751 RNA interference 0.036900464

CPG GO:0007274 0.017190494 neuromuscular synaptic transmission 0.038720046

CPG GO:0061484 0.017221633 hematopoietic stem cell homeostasis 0.038720046

CPG GO:0033081 0.017221633 regulation of T cell differentiation in thymus 0.038720046

CPG GO:0051300 0.017221633 spindle pole body organization 0.038720046

CPG GO:0007421 0.017221633 stomatogastric nervous system development 0.038720046

CPG GO:0034116 0.017221633 positive regulation of heterotypic cell-cell adhesion 0.038720046 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0070262 0.017221633 peptidyl-serine dephosphorylation 0.038720046

CHH GO:0007204 0.017635281 positive regulation of cytosolic calcium ion concentration 0.040659623

CPG GO:0006897 0.017654527 endocytosis 0.038720046

CHH GO:0045666 0.017717994 positive regulation of neuron differentiation 0.040659623

197 CHG GO:0010454 0.017937861 negative regulation of cell fate commitment 0.036900464

CHG GO:0051642 0.017937861 centrosome localization 0.036900464

CHH GO:0060059 0.01796588 embryonic retina morphogenesis in camera-type eye 0.040659623

CHH GO:1902630 0.01796588 regulation of membrane hyperpolarization 0.040659623

CHH GO:0006564 0.01796588 L-serine biosynthetic process 0.040659623

CHH GO:0002087 0.01796588 regulation of respiratory gaseous exchange by neurological system process 0.040659623

CHH GO:1903351 0.01796588 cellular response to dopamine 0.040659623

CHH GO:0007191 0.01796588 adenylate cyclase-activating dopamine receptor signaling pathway 0.040659623 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0099509 0.01796588 regulation of presynaptic cytosolic calcium ion concentration 0.040659623

CHH GO:1990834 0.01796588 response to odorant 0.040659623

CHH GO:1904016 0.01796588 response to Thyroglobulin triiodothyronine 0.040659623

CHH GO:0035261 0.01796588 external genitalia morphogenesis 0.040659623

198 CHH GO:0034116 0.01796588 positive regulation of heterotypic cell-cell adhesion 0.040659623

CHH GO:1904659 0.01796588 glucose transmembrane transport 0.040659623

CHH GO:2000311 0.01796588 regulation of AMPA receptor activity 0.040659623

CPG GO:0050803 0.018194022 regulation of synapse structure or activity 0.038720046

CHH GO:0016331 0.018232402 morphogenesis of embryonic epithelium 0.040809841

CHH GO:0045087 0.018411882 innate immune response 0.040809841

CPG GO:0009611 0.018594327 response to wounding 0.038720046

CPG GO:0097480 0.018693919 establishment of synaptic vesicle localization 0.038720046 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0099003 0.018693919 vesicle-mediated transport in synapse 0.038720046

CHH GO:0051241 0.018790097 negative regulation of multicellular organismal process 0.040894957

CHG GO:0046928 0.018926099 regulation of neurotransmitter secretion 0.036900464

CHH GO:0007444 0.019134629 imaginal disc development 0.040894957

199 CPG GO:0001751 0.019201441 compound eye photoreceptor cell differentiation 0.038720046

CHH GO:0048563 0.019885386 post-embryonic animal organ morphogenesis 0.040894957

CPG GO:0048859 0.020090091 formation of anatomical boundary 0.038720046

CPG GO:0031960 0.020090091 response to corticosteroid 0.038720046

CPG GO:0035295 0.020105784 tube development 0.038720046

CPG GO:0010770 0.020272148 positive regulation of cell morphogenesis involved in differentiation 0.038720046

CPG GO:0007411 0.020344481 axon guidance 0.038720046

CPG GO:0044700 0.02045535 single organism signaling 0.038720046 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0098609 0.020648864 cell-cell adhesion 0.040894957

CHG GO:0014016 0.020899966 neuroblast differentiation 0.036900464

CHG GO:0035071 0.020899966 salivary gland cell autophagic cell death 0.036900464

CPG GO:0048869 0.020960838 cellular developmental process 0.038720046

200 CPG GO:0032501 0.021465316 multicellular organismal process 0.038720046

CHH GO:0072507 0.021493475 divalent inorganic cation homeostasis 0.040894957

CHH GO:0042330 0.021522882 taxis 0.040894957

CHG GO:0000027 0.021885596 ribosomal large subunit assembly 0.036900464

CHH GO:0002164 0.022066154 larval development 0.040894957

CHH GO:0010769 0.022066154 regulation of cell morphogenesis involved in differentiation 0.040894957

CHH GO:0060322 0.022149193 head development 0.040894957

CHH GO:0048806 0.022854713 genitalia development 0.040894957 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0045746 0.022870358 negative regulation of Notch signaling pathway 0.036900464

CHG GO:0017158 0.022870358 regulation of calcium ion-dependent exocytosis 0.036900464

CPG GO:0045498 0.02289744 sex comb development 0.038720046

CPG GO:0006531 0.02289744 aspartate metabolic process 0.038720046

201 CPG GO:0006616 0.02289744 SRP-dependent cotranslational protein targeting to membrane, translocation 0.038720046

CPG GO:0035385 0.02289744 Roundabout signaling pathway 0.038720046

CPG GO:0016200 0.02289744 synaptic target attraction 0.038720046

CPG GO:0035157 0.02289744 negative regulation of fusion cell fate speciﬁcation 0.038720046

CPG GO:0046929 0.02289744 negative regulation of neurotransmitter secretion 0.038720046

CPG GO:0048645 0.023179347 animal organ formation 0.038720046

CPG GO:0036465 0.023179347 synaptic vesicle recycling 0.038720046

CHH GO:0007166 0.023666191 cell surface receptor signaling pathway 0.040894957 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0050767 0.023686925 regulation of neurogenesis 0.036900464

CHG GO:0008347 0.023854252 glial cell migration 0.036900464

CHH GO:2000302 0.023883976 positive regulation of synaptic vesicle exocytosis 0.040894957

CHH GO:0061002 0.023883976 negative regulation of dendritic spine morphogenesis 0.040894957

202 CHH GO:0072340 0.023883976 cellular lactam catabolic process 0.040894957

CHH GO:0043576 0.023883976 regulation of respiratory gaseous exchange 0.040894957

CHH GO:0006013 0.023883976 mannose metabolic process 0.040894957

CHH GO:0097119 0.023883976 postsynaptic density protein 95 clustering 0.040894957

CHH GO:0035385 0.023883976 Roundabout signaling pathway 0.040894957

CHH GO:0016183 0.023883976 synaptic vesicle coating 0.040894957

CHH GO:0016200 0.023883976 synaptic target attraction 0.040894957

CHH GO:1904861 0.023883976 excitatory synapse assembly 0.040894957 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:1904886 0.023883976 beta-catenin destruction complex disassembly 0.040894957

CHH GO:0071869 0.023883976 response to catecholamine 0.040894957

CHH GO:0071880 0.023883976 adenylate cyclase-activating adrenergic receptor signaling pathway 0.040894957

CHH GO:0071542 0.023883976 dopaminergic neuron differentiation 0.040894957

203 CHH GO:1900409 0.023883976 positive regulation of cellular response to oxidative stress 0.040894957

CHH GO:1900244 0.023883976 positive regulation of synaptic vesicle endocytosis 0.040894957

CHH GO:0048259 0.023966347 regulation of receptor-mediated endocytosis 0.040894957

CHH GO:0010977 0.023966347 negative regulation of neuron projection development 0.040894957

CPG GO:0051961 0.024113283 negative regulation of nervous system development 0.039827558

CPG GO:0030030 0.024517597 cell projection organization 0.040045409

CHH GO:0060560 0.025054754 developmental growth involved in morphogenesis 0.041833114

CHH GO:0021953 0.025099869 central nervous system neuron differentiation 0.041833114 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0007616 0.025099869 long-term memory 0.041833114

CPG GO:0007399 0.025177509 nervous system development 0.040490043

CPG GO:0022416 0.025340707 chaeta development 0.040490043

CHG GO:0048864 0.025819439 stem cell development 0.036900464

204 CHG GO:0048865 0.025819439 stem cell fate commitment 0.036900464

CHG GO:1903429 0.025819439 regulation of cell maturation 0.036900464

CHH GO:0048732 0.026501808 gland development 0.042497427

CHG GO:0060033 0.026800734 anatomical structure regression 0.036900464

CHG GO:0060581 0.026800734 cell fate commitment involved in pattern speciﬁcation 0.036900464

CHG GO:0042659 0.026800734 regulation of cell fate speciﬁcation 0.036900464

CPG GO:0007275 0.027042039 multicellular organism development 0.040990531

CPG GO:0098609 0.027045681 cell-cell adhesion 0.040990531 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0071822 0.027073902 protein complex subunit organization 0.042497427

CPG GO:0050769 0.027110158 positive regulation of neurogenesis 0.040990531

CPG GO:0032990 0.027187055 cell part morphogenesis 0.040990531

CHG GO:0048513 0.027333677 animal organ development 0.036900464

205 CHH GO:0055082 0.02741169 cellular chemical homeostasis 0.042497427

CHH GO:0097485 0.028340246 neuron projection guidance 0.042497427

CPG GO:2000114 0.028541177 regulation of establishment of cell polarity 0.040990531

CPG GO:0048499 0.028541177 synaptic vesicle membrane organization 0.040990531

CPG GO:0019229 0.028541177 regulation of vasoconstriction 0.040990531

CPG GO:0014068 0.028541177 positive regulation of phosphatidylinositol 3-kinase signaling 0.040990531

CPG GO:0046845 0.028541177 branched duct epithelial cell fate determination, open tracheal system 0.040990531

CPG GO:0030890 0.028541177 positive regulation of B cell proliferation 0.040990531 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0048468 0.028615857 cell development 0.042497427

CHG GO:0043954 0.02876073 cellular component maintenance 0.036978081

CHG GO:0010160 0.02876073 formation of animal organ boundary 0.036978081

CHH GO:0043434 0.028916386 response to peptide hormone 0.042497427

206 CPG GO:0097305 0.029000103 response to alcohol 0.040990531

CPG GO:0040017 0.029000103 positive regulation of locomotion 0.040990531

CHH GO:0048858 0.029134979 cell projection morphogenesis 0.042497427

CHH GO:0050774 0.02916879 negative regulation of dendrite morphogenesis 0.042497427

CHG GO:0042551 0.029739432 neuron maturation 0.037347193

CHH GO:0048047 0.029767147 mating behavior, sex discrimination 0.042497427

CHH GO:0005984 0.029767147 disaccharide metabolic process 0.042497427

CHH GO:0048499 0.029767147 synaptic vesicle membrane organization 0.042497427 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0021794 0.029767147 thalamus development 0.042497427

CHH GO:0048268 0.029767147 clathrin coat assembly 0.042497427

CHH GO:0097120 0.029767147 receptor localization to synapse 0.042497427

CHH GO:0048789 0.029767147 cytoskeletal matrix organization at active zone 0.042497427

207 CHH GO:0071353 0.029767147 cellular response to interleukin-4 0.042497427

CHH GO:0071346 0.029767147 cellular response to interferon-gamma 0.042497427

CHH GO:0071474 0.029767147 cellular hyperosmotic response 0.042497427

CHH GO:0071868 0.029767147 cellular response to monoamine stimulus 0.042497427

CHH GO:1901642 0.029767147 nucleoside transmembrane transport 0.042497427

CHH GO:0000380 0.02984703 alternative mRNA splicing, via spliceosome 0.042497427

CHH GO:0045807 0.02984703 positive regulation of endocytosis 0.042497427

CPG GO:0034110 0.02989864 regulation of homotypic cell-cell adhesion 0.041858095 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0048736 0.030743761 appendage development 0.043474802

CHH GO:0007155 0.030937882 cell adhesion 0.043474802

CPG GO:0008283 0.031446109 cell proliferation 0.04328012

CHG GO:0035050 0.032670362 embryonic heart tube development 0.039204434

208 CHG GO:0008584 0.032670362 male gonad development 0.039204434

CHH GO:0048675 0.033623161 axon extension 0.0437563

CHH GO:0008045 0.033623161 motor neuron axon guidance 0.0437563

CHH GO:0030336 0.033623161 negative regulation of cell migration 0.0437563

CHH GO:0030431 0.033623161 sleep 0.0437563

CHH GO:0009791 0.033733118 post-embryonic development 0.0437563

CPG GO:0032012 0.03415302 regulation of ARF protein signal transduction 0.04328012

CPG GO:0097320 0.03415302 plasma membrane tubulation 0.04328012 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0048790 0.03415302 maintenance of presynaptic active zone structure 0.04328012

CPG GO:0007438 0.03415302 oenocyte development 0.04328012

CPG GO:0007402 0.03415302 ganglion mother cell fate determination 0.04328012

CPG GO:0010560 0.03415302 positive regulation of glycoprotein biosynthetic process 0.04328012

209 CPG GO:0046533 0.03415302 negative regulation of photoreceptor cell differentiation 0.04328012

CPG GO:0009435 0.03415302 NAD biosynthetic process 0.04328012

CPG GO:1900026 0.03415302 positive regulation of substrate adhesion-dependent cell spreading 0.04328012

CPG GO:0071549 0.03415302 cellular response to dexamethasone stimulus 0.04328012

CPG GO:0022603 0.034481852 regulation of anatomical structure morphogenesis 0.043323352

CHH GO:0001754 0.03453045 eye photoreceptor cell differentiation 0.0437563

CHH GO:0051960 0.034669218 regulation of nervous system development 0.0437563

CHH GO:0048521 0.034888853 negative regulation of behavior 0.0437563 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0001941 0.034932066 postsynaptic membrane organization 0.0437563

CPG GO:0050804 0.035081229 modulation of synaptic transmission 0.043557103

CPG GO:0040011 0.035260512 locomotion 0.043557103

CHG GO:0007405 0.035593542 neuroblast proliferation 0.041783723

210 CHH GO:0030917 0.035615593 midbrain-hindbrain boundary development 0.0437563

CHH GO:2000234 0.035615593 positive regulation of rRNA processing 0.0437563

CHH GO:0032230 0.035615593 positive regulation of synaptic transmission, GABAergic 0.0437563

CHH GO:0048790 0.035615593 maintenance of presynaptic active zone structure 0.0437563

CHH GO:0014029 0.035615593 neural crest formation 0.0437563

CHH GO:0006691 0.035615593 leukotriene metabolic process 0.0437563

CHH GO:0099084 0.035615593 postsynaptic specialization organization 0.0437563

CHH GO:0051590 0.035615593 positive regulation of neurotransmitter transport 0.0437563 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0007301 0.035615593 female germline ring canal formation 0.0437563

CHH GO:0051968 0.035615593 positive regulation of synaptic transmission, glutamatergic 0.0437563

CHH GO:0035204 0.035615593 negative regulation of lamellocyte differentiation 0.0437563

CHH GO:0034121 0.035615593 regulation of toll-like receptor signaling pathway 0.0437563

211 CHH GO:0030723 0.035615593 ovarian fusome organization 0.0437563

CPG GO:0072583 0.036016527 clathrin-dependent endocytosis 0.044120245

CPG GO:0007154 0.037412289 cell communication 0.044872278

CHH GO:0007568 0.037535367 aging 0.045643044

CHH GO:0007156 0.037575902 homophilic cell adhesion via plasma membrane adhesion molecules 0.045643044

CHG GO:0017145 0.03947909 stem cell division 0.045358954

CPG GO:0032026 0.039733145 response to magnesium ion 0.044872278

CPG GO:0048058 0.039733145 compound eye corneal lens development 0.044872278 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0048521 0.039733145 negative regulation of behavior 0.044872278

positive regulation of adenylate cyclase activity involved in CPG GO:0010579 0.039733145 0.044872278 G-protein coupled receptor signaling pathway

CPG GO:0035155 0.039733145 negative regulation of terminal cell fate speciﬁcation, open tracheal system 0.044872278

CPG GO:0035225 0.039733145 determination of genital disc primordium 0.044872278 212

CPG GO:0046673 0.039733145 negative regulation of compound eye retinal cell programmed cell death 0.044872278

CPG GO:0034260 0.039733145 negative regulation of GTPase activity 0.044872278

CPG GO:1900242 0.039733145 regulation of synaptic vesicle endocytosis 0.044872278

CPG GO:0048513 0.03998822 animal organ development 0.044872278

CHH GO:0007622 0.040305575 rhythmic behavior 0.046899783

CHG GO:0061326 0.040448334 renal tubule development 0.045504376

CPG GO:0035150 0.041228398 regulation of tube size 0.045913443 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0048058 0.041429515 compound eye corneal lens development 0.046899783

CHH GO:0006000 0.041429515 fructose metabolic process 0.046899783

CHH GO:0019852 0.041429515 L-ascorbic acid metabolic process 0.046899783

CHH GO:0006828 0.041429515 manganese ion transport 0.046899783

213 CHH GO:0045613 0.041429515 regulation of plasmatocyte differentiation 0.046899783

CHH GO:0035155 0.041429515 negative regulation of terminal cell fate speciﬁcation, open tracheal system 0.046899783

CHH GO:0003344 0.041429515 pericardium morphogenesis 0.046899783

CHH GO:0016999 0.041429515 antibiotic metabolic process 0.046899783

CHH GO:0071732 0.041429515 cellular response to nitric oxide 0.046899783

CHH GO:0070142 0.041429515 synaptic vesicle budding 0.046899783

CHH GO:0010243 0.041544761 response to organonitrogen compound 0.046899783

CHH GO:0042051 0.041567532 compound eye photoreceptor development 0.046899783 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0051271 0.041697971 negative regulation of cellular component movement 0.046899783

CPG GO:0030707 0.04177326 ovarian follicle cell development 0.046170445

CHH GO:0042221 0.041882597 response to chemical 0.046899783

CHH GO:0051705 0.042287405 multi-organism behavior 0.047107731

214 CHH GO:0050773 0.043108393 regulation of dendrite development 0.047529767

CHH GO:0008593 0.043108393 regulation of Notch signaling pathway 0.047529767

CHG GO:0035051 0.043350934 cardiocyte differentiation 0.047774498

CPG GO:0007165 0.044067402 signal transduction 0.047240959

CHG GO:0045807 0.044316758 positive regulation of endocytosis 0.047862099

CPG GO:0061331 0.045281729 epithelial cell proliferation involved in Malpighian tubule morphogenesis 0.047240959

CPG GO:0055070 0.045281729 copper ion homeostasis 0.047240959

CPG GO:0021772 0.045281729 olfactory bulb development 0.047240959 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CPG GO:0097756 0.045281729 negative regulation of blood vessel diameter 0.047240959

CPG GO:0021903 0.045281729 rostrocaudal neural tube patterning 0.047240959

CPG GO:0016330 0.045281729 second mitotic wave involved in compound eye morphogenesis 0.047240959

CHG GO:0016079 0.045281729 synaptic vesicle exocytosis 0.04794536

215 CPG GO:0007298 0.045312757 border follicle cell migration 0.047240959

CHH GO:0044700 0.045604476 single organism signaling 0.048104073

CHH GO:2000370 0.046268606 positive regulation of clathrin-dependent endocytosis 0.048104073

CPG GO:0000122 0.046415435 negative regulation of transcription from RNA polymerase II promoter 0.047538869

CPG GO:0007010 0.046533076 cytoskeleton organization 0.047538869

CPG GO:0051648 0.046568688 vesicle localization 0.047538869

CHG GO:0009888 0.046602299 tissue development 0.048099852

CHH GO:0048585 0.046771532 negative regulation of response to stimulus 0.048104073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHG GO:0048675 0.047209114 axon extension 0.048099852

CHH GO:2000463 0.047209114 positive regulation of excitatory postsynaptic potential 0.048104073

CHH GO:0060325 0.047209114 face morphogenesis 0.048104073

CHH GO:0021772 0.047209114 olfactory bulb development 0.048104073

216 CHH GO:0097104 0.047209114 postsynaptic membrane assembly 0.048104073

CHH GO:0045314 0.047209114 regulation of compound eye photoreceptor development 0.048104073

CHH GO:0006554 0.047209114 lysine catabolic process 0.048104073

CHH GO:0045611 0.047209114 negative regulation of hemocyte differentiation 0.048104073

CHH GO:1903421 0.047209114 regulation of synaptic vesicle recycling 0.048104073

CHH GO:0016080 0.047209114 synaptic vesicle targeting 0.048104073

CHH GO:0050667 0.047209114 homocysteine metabolic process 0.048104073

CHH GO:0007205 0.047209114 protein kinase C-activating G-protein coupled receptor signaling pathway 0.048104073 Table A.2 continued from previous page

Genomic Gene ontology P-value Biological processes FDR Context

CHH GO:0017085 0.047209114 response to insecticide 0.048104073

CHH GO:1905950 0.047209114 monosaccharide transmembrane transport 0.048104073

CPG GO:0051240 0.047497996 positive regulation of multicellular organismal process 0.048153141

CHH GO:0006898 0.048061387 receptor-mediated endocytosis 0.048741501

217 CPG GO:0001708 0.048115822 cell fate speciﬁcation 0.048225721

CHG GO:0035295 0.048115923 tube development 0.048115923

CPG GO:0007163 0.048225721 establishment or maintenance of cell polarity 0.048225721

CHH GO:0008340 0.048925219 determination of adult lifespan 0.049384611

CHH GO:0043436 0.049566043 oxoacid metabolic process 0.04979766

CHH GO:0007417 0.049816239 central nervous system development 0.049816239 TABLE A.3: These tables show genes differentially expressed between treated and untreated bees with name of transcript for each gene.

LOC log2FoldChange pvalue padj name

LOC100651433 -2.000086634 8.63E-18 2.77E-14 protein lethal(2)essential for life

sushi, von Willebrand factor type A, EGF and pentraxin LOC100648516 -1.736812384 0.001866387 0.048440531 domain-containing

LOC100643857 -1.706166699 0.00179133 0.047127652 caspase-1

LOC105666599 -1.449141716 1.64E-12 1.98E-09 signal transducer and activator of transcription C-like

218 LOC100651530 -1.429372048 9.19E-08 2.53E-05 uncharacterized LOC100651530

LOC100631082 -1.350298074 3.94E-06 0.000555066 homeotic protein antennapedia

LOC100646094 -1.295140373 1.40E-11 1.13E-08 peptide methionine sulfoxide reductase

LOC100643859 -1.212128658 0.001049207 0.032380813 anoctamin-4

LOC110119550 -1.204163034 0.000949385 0.030532976 glycine-rich cell wall structural protein 1.0-like

LOC100643972 -1.183464157 2.01E-15 3.23E-12 proton-coupled amino acid transporter-like protein pathetic

LOC100651391 -1.0685149 0.000700338 0.025182455 facilitated trehalose transporter Tret1-like

LOC100644995 -1.057546993 1.02E-07 2.73E-05 uncharacterized LOC100644995 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100643224 -1.00903565 0.000235373 0.012591147 homeotic protein Sex combs reduced

LOC100642982 -1.00680295 0.00024189 0.012800912 homeotic protein deformed

LOC100651204 -0.962513941 9.15E-11 5.87E-08 protein glass

LOC100644150 -0.875223238 0.000589668 0.023080948 protein scarlet

LOC100642642 -0.862877011 0.001714335 0.045981432 uncharacterized LOC100642642

219 LOC100649441 -0.809777164 1.12E-05 0.001283457 probable cytochrome P450 28d1

LOC100646889 -0.803400665 0.000156074 0.009572225 collagen alpha-1(IX) chain

LOC100644559 -0.769433852 0.000462165 0.019778625 uncharacterized LOC100644559

LOC100648756 -0.748402397 4.16E-09 2.11E-06 uncharacterized LOC100648756

LOC100651542 -0.734871323 0.000230507 0.01246939 dendritic arbor reduction protein 1

LOC110119834 -0.719403784 9.54E-05 0.006855111 uncharacterized LOC110119834

LOC105666209 -0.715342278 0.001455085 0.040611623 protein PTHB1

LOC105666953 -0.714728785 0.001759884 0.046811951 uncharacterized LOC105666953 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC105666619 -0.70884055 0.001005652 0.031501354 uncharacterized LOC105666619

LOC100647528 -0.701029131 1.71E-05 0.001771765 neither inactivation nor afterpotential protein C

LOC100651053 -0.700662185 0.001182838 0.034937273 hyaluronidase

LOC100647949 -0.690730077 0.000158413 0.009591261 uncharacterized LOC100647949

LOC100648970 -0.666295658 3.19E-07 6.83E-05 uncharacterized LOC100648970

220 LOC100650672 -0.658088255 1.37E-07 3.47E-05 uncharacterized LOC100650672

LOC100644143 -0.639081523 0.000113941 0.007672261 putative inorganic phosphate cotransporter

LOC100645006 -0.635356737 2.17E-05 0.002147255 uncharacterized LOC100645006

LOC100647268 -0.619369255 0.000122781 0.007988212 matrix metalloproteinase-14

LOC100649302 -0.613721334 0.000449052 0.01942121 sensory neuron membrane protein 1

LOC100643639 -0.609747716 0.00045196 0.019428215 histone H1A, sperm

LOC100649106 -0.589243396 4.73E-11 3.26E-08 putative glutathione-speciﬁc gamma-glutamylcyclotransferase 2

LOC100646667 -0.584175548 0.001098447 0.033156582 circadian clock-controlled protein Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100647771 -0.580202914 2.19E-05 0.002147255 uncharacterized LOC100647771

LOC110120070 -0.576910031 2.97E-12 3.17E-09 leucine-rich repeat-containing protein 15-like

LOC100644036 -0.55645317 1.47E-07 3.63E-05 protein Skeletor, isoforms B/C

LOC100652167 -0.554752811 5.77E-08 1.74E-05 monocarboxylate transporter 10

LOC100648765 -0.553225804 9.17E-05 0.006637895 endothelin-converting enzyme homolog

221 LOC100649129 -0.499624549 0.001014168 0.031501354 uncharacterized LOC100649129

LOC100644864 -0.493011684 2.52E-06 0.000397741 carboxypeptidase Q

LOC100649077 -0.483395636 1.89E-06 0.000319378 N66 matrix protein

LOC105666952 -0.471109398 6.12E-06 0.000785905 uncharacterized LOC105666952

LOC100642770 -0.469886718 3.45E-14 4.75E-11 eukaryotic translation initiation factor 2 subunit 2

LOC100648256 -0.469777492 0.000361458 0.016977953 integrin beta-nu

LOC100650309 -0.46387617 1.53E-06 0.000267602 cytochrome P450 9e2

LOC100651177 -0.462084712 0.00043418 0.019068382 serine proteinase stubble Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100650142 -0.453665341 1.24E-06 0.000220479 furin-like protease 1

LOC100646517 -0.451519589 1.98E-06 0.000328624 selenium-binding protein 1-B

LOC100650710 -0.447667769 0.000109365 0.007481114 general odorant-binding protein 71

LOC100651917 -0.445678316 0.001263963 0.036493644 organic cation transporter 1

LOC100646870 -0.441670717 0.000104374 0.007282752 uncharacterized LOC100646870

222 LOC100643518 -0.440577557 2.14E-05 0.002147255 histidine decarboxylase

LOC100644989 -0.438904339 0.001593669 0.043348695 thyroid transcription factor 1-associated protein 26

LOC100645609 -0.427657946 2.49E-07 5.85E-05 large neutral amino acids transporter small subunit 1

LOC100647133 -0.423475219 0.001131488 0.033835718 myosin-IIIb

LOC100649408 -0.423035347 0.000169079 0.010112172 protein CBFA2T2

LOC100643624 -0.421334869 0.00168507 0.045577366 sorbitol dehydrogenase

LOC100647937 -0.419793355 3.27E-05 0.002943845 uncharacterized LOC100647937

LOC100642602 -0.418739126 2.83E-07 6.34E-05 inorganic pyrophosphatase Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100649579 -0.416120404 0.001025539 0.031752149 L-lactate dehydrogenase

LOC110119518 -0.412229834 0.000703014 0.025182455 uncharacterized LOC110119518

LOC100648651 -0.405794117 0.000539242 0.021545059 scavenger receptor class B member 1

LOC100651810 -0.403436476 0.000305162 0.015304191 peroxisomal leader peptide-processing protease

LOC100647601 -0.401127074 0.000862003 0.029123604 facilitated trehalose transporter Tret1

223 LOC110119814 -0.400489218 0.000115562 0.007698432 uncharacterized LOC110119814

LOC100648070 -0.400263404 0.000957151 0.030619281 enhancer of split mbeta protein-like

LOC100646242 -0.39895362 9.83E-05 0.006988246 leucine-rich repeat-containing protein 15

LOC100648157 -0.389710005 8.59E-08 2.43E-05 programmed cell death protein 6

LOC100649380 -0.388837422 0.001154196 0.034331728 nucleolar protein 12

LOC110119287 -0.383441522 8.27E-12 7.24E-09 neutral and basic amino acid transport protein rBAT

LOC100645983 -0.383133106 1.40E-08 5.37E-06 large neutral amino acids transporter small subunit 2

LOC100644078 -0.369245606 0.000321122 0.015854657 uncharacterized LOC100644078 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100652102 -0.365949169 1.32E-05 0.001439564 28S ribosomal protein S17, mitochondrial

LOC100644110 -0.36029153 6.11E-06 0.000785905 uncharacterized LOC100644110

LOC110119520 -0.354957634 0.000258258 0.013231834 uncharacterized LOC110119520

LOC100648549 -0.354232524 2.97E-08 9.95E-06 cytochrome c

LOC100642958 -0.352787367 0.000181817 0.010546494 uncharacterized LOC100642958

224 LOC100643720 -0.348589733 0.000885265 0.02949671 thrombospondin type-1 domain-containing protein 4

LOC100647411 -0.343109778 0.001010204 0.031501354 60S ribosomal protein L11

LOC100652040 -0.342556866 0.00027392 0.013881961 calpain-C

LOC105666762 -0.34140382 0.000119708 0.007841297 bone morphogenetic protein 4

LOC100647883 -0.340395955 3.14E-06 0.000480495 UDP-glucuronosyltransferase 2B15

LOC100642474 -0.338283847 2.34E-06 0.000375382 protein giant-lens

LOC100648227 -0.332924085 0.000481413 0.020331268 uncharacterized LOC100648227

LOC100650109 -0.328836073 3.29E-05 0.002943845 bicaudal D-related protein homolog Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100646731 -0.325230983 0.001304548 0.037164173 RNA-binding protein 8A

LOC100643860 -0.32152883 2.15E-05 0.002147255 uncharacterized LOC100643860

LOC100642767 -0.319526586 9.95E-05 0.006992246 Krueppel-like factor 10

LOC100646310 -0.315407374 0.001277221 0.036493644 transducin beta-like protein 2

LOC100643981 -0.312735456 4.94E-05 0.003960001 high-afﬁnity choline transporter 1

225 LOC100646922 -0.312177602 8.91E-05 0.006497079 receptor-type guanylate cyclase gcy-3

LOC100651362 -0.304800399 0.000341292 0.016268835 calumenin-B

LOC100650498 -0.304692509 0.0006008 0.02323336 39S ribosomal protein L53, mitochondrial

LOC100648258 -0.304663228 0.00033818 0.016268835 serine/threonine-protein kinase SBK1

LOC100644817 -0.304448492 0.000330773 0.016085909 60 kDa heat shock protein, mitochondrial

LOC100647489 -0.303342019 0.000254335 0.013231834 sodium bicarbonate cotransporter 3

LOC100647117 -0.30317075 0.00031765 0.015766225 inward rectiﬁer potassium channel 4

LOC100650877 -0.300685287 0.000133765 0.008473861 protein CNPPD1 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100644399 -0.298803816 8.52E-05 0.006317176 unc-112-related protein

LOC100651452 -0.298777046 0.000110325 0.007481114 cytochrome b-c1 complex subunit 7

LOC100646899 -0.298270813 2.79E-05 0.002634207 cysteine-rich protein 1

LOC100648183 -0.293309451 0.000826909 0.028305725 enhancer of split m7 protein

LOC100644714 -0.293062004 0.000258343 0.013231834 nuclear cap-binding protein subunit 2

226 complement component 1 Q subcomponent-binding protein, LOC100649532 -0.286258752 1.97E-05 0.002018891 mitochondrial

LOC100650321 -0.284679576 0.00031463 0.015697285 zinc ﬁnger matrin-type protein 2

LOC100645110 -0.281197913 0.000998674 0.031501354 microﬁbrillar-associated protein 1

LOC100648341 -0.280665021 3.30E-05 0.002943845 mitochondrial import inner membrane translocase subunit Tim16

LOC100648104 -0.280169457 0.000496777 0.020797681 protein tyrosine phosphatase type IVA 1

LOC100644923 -0.279687382 0.000255196 0.013231834 uncharacterized LOC100644923

LOC100644676 -0.278551522 8.18E-06 0.000996687 E3 ubiquitin-protein ligase MYLIP Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100646126 -0.274419066 0.001721666 0.046014968 glutamate decarboxylase 2

LOC100649542 -0.272394684 7.33E-05 0.005555754 G-protein coupled receptor moody

LOC100648892 -0.271976196 0.000608348 0.023431128 quinone oxidoreductase

LOC100651889 -0.271372257 0.000404652 0.018466301 MAP/microtubule afﬁnity-regulating kinase 3

LOC100645422 -0.269974504 0.001725143 0.046014968 reversion-inducing cysteine-rich protein with Kazal motifs

227 LOC100642544 -0.2693764 4.59E-08 1.43E-05 uncharacterized LOC100642544

LOC100643725 -0.269119349 0.000192359 0.011091173 septum formation protein Maf

LOC100649320 -0.261743416 0.000131692 0.00839778 Na(+)/H(+) exchange regulatory cofactor NHE-RF1

LOC100646771 -0.261423552 4.33E-06 0.000587696 protein pelota

LOC100646091 -0.257475271 0.000159182 0.009591261 cilia- and ﬂagella-associated protein 91

LOC100649276 -0.255704742 0.000149393 0.00928065 activator of 90 kDa heat shock protein ATPase homolog 1

LOC100647094 -0.253081733 1.05E-05 0.001245367 zinc transporter 1

LOC100649074 -0.249020787 0.000731199 0.025710557 nuclear transcription factor Y subunit gamma Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100647360 -0.247732094 0.000657689 0.024357243 patched domain-containing protein 3

LOC100643466 -0.247494728 0.000635383 0.024076341 grpE protein homolog, mitochondrial

LOC110119145 -0.244929472 1.03E-05 0.00123476 ribosome biogenesis protein NSA2 homolog

LOC100647310 -0.243675463 0.000726768 0.025710557 bumetanide-sensitive sodium-(potassium)-chloride cotransporter

LOC100645971 -0.239254384 0.000519199 0.021213876 low-density lipoprotein receptor-related protein 2

228 LOC110119386 -0.239215597 0.000210766 0.011850383 nesprin-1-like

LOC100643607 -0.237900947 0.00021994 0.012224224 carbonic anhydrase 1

LOC100642487 -0.236980954 0.000141536 0.008849658 G kinase-anchoring protein 1

LOC100650895 -0.236926822 0.000554472 0.021971234 reactive oxygen species modulator 1

LOC100646350 -0.236864054 0.000530282 0.021314962 translocon-associated protein subunit alpha

LOC100645809 -0.236631033 3.87E-06 0.000555066 aspartate–tRNA ligase, cytoplasmic

LOC100650794 -0.234751926 0.000257055 0.013231834 proﬁlin

LOC100648541 -0.232272102 0.000425118 0.019068382 F-box only protein 44 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100648961 -0.232218965 0.000514064 0.021213876 uncharacterized LOC100648961

LOC110119804 -0.230620582 0.000514507 0.021213876 uncharacterized LOC110119804

LOC100644971 -0.230408111 0.000519937 0.021213876 elongation factor Ts, mitochondrial

LOC100649164 -0.228239367 0.000118466 0.007813101 elongation factor 1-beta’

LOC100649609 -0.227783515 0.001199257 0.035206251 mitochondrial import inner membrane translocase subunit TIM50-C

229 LOC100648817 -0.224219166 0.001243624 0.036068838 protein BUD31 homolog

LOC100643135 -0.221597777 2.66E-06 0.00041334 protein l(2)37Cc

LOC100644191 -0.221511744 0.000434576 0.019068382 serine/arginine-rich splicing factor 2

LOC100646420 -0.221490931 4.19E-05 0.003536691 neurexin-4

LOC100642576 -0.221007281 0.000241953 0.012800912 nuclear migration protein nudC

LOC100646214 -0.220916152 0.000340724 0.016268835 uncharacterized LOC100646214

LOC100644325 -0.220271264 0.000674624 0.024793714 small nuclear ribonucleoprotein-associated protein B

LOC100652239 -0.220163454 0.001076151 0.033106254 transcriptional activator protein Pur-beta Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100650356 -0.219040569 0.0008853 0.02949671 SUMO-activating enzyme subunit 1

LOC100652188 -0.218115857 0.000686921 0.024959872 copper chaperone for superoxide dismutase

LOC100642779 -0.21798145 6.70E-07 0.000130499 G protein-coupled receptor kinase 2

LOC100650583 -0.216597684 0.001770834 0.046822728 general transcription factor 3C polypeptide 1

LOC100646553 -0.213789423 0.001492938 0.041308912 NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 7

230 LOC100647892 -0.211062182 0.0006858 0.024959872 neuroﬁlament heavy polypeptide

LOC100651427 -0.209516748 2.33E-06 0.000375382 S-phase kinase-associated protein 1

LOC100643654 -0.209167333 0.000908922 0.029768749 uncharacterized family 31 glucosidase KIAA1161

LOC100644830 -0.208870715 0.001962066 0.049980775 cleavage stimulation factor subunit 1

LOC100645142 -0.208470496 0.000115928 0.007698432 translation initiation factor eIF-2B subunit delta

LOC100651878 -0.207726133 0.001893574 0.048957656 WD repeat-containing protein 55 homolog

LOC100649719 -0.20744295 0.00047427 0.020117841 glutamate–cysteine ligase regulatory subunit

LOC100648832 -0.205952868 5.26E-05 0.004185717 uncharacterized LOC100648832 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100650517 -0.203505912 1.62E-05 0.001736432 nitric oxide synthase, salivary gland

LOC105665676 -0.202711907 0.001446086 0.040595803 chaoptin

LOC100644288 -0.201594403 0.000627469 0.023975779 papilin

LOC100647102 -0.198224509 0.000174869 0.010267185 DNA-directed RNA polymerase I subunit RPA12

LOC100649691 -0.196642751 0.000884183 0.02949671 glycerol-3-phosphate phosphatase

231 LOC100643867 -0.195404677 3.78E-05 0.003262415 ribonuclease Z, mitochondrial

LOC100643603 -0.193796591 0.000896651 0.029652175 uncharacterized LOC100643603

LOC100649255 -0.191904518 0.000446982 0.01942121 protein unc-79 homolog

LOC105667138 -0.191850856 0.001108422 0.033353123 acyl carrier protein, mitochondrial

LOC100648998 -0.189669443 0.000434676 0.019068382 Golgi apparatus membrane protein-like protein CG50

LOC100643070 -0.187997006 0.000173084 0.010267185 ryanodine receptor

LOC100650375 -0.187954704 0.000105342 0.007297406 integrator complex subunit 2

LOC100645135 -0.185946316 0.000935499 0.030432162 uncharacterized LOC100645135 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100645196 -0.18362644 0.000388348 0.017935807 protein bric-a-brac 1

LOC100650622 -0.183317147 0.000895952 0.029652175 geranylgeranyl transferase type-2 subunit beta

LOC100648608 -0.181349045 0.000918708 0.029987254 40S ribosomal protein S23

LOC100650315 -0.18130178 0.000745063 0.025899671 division abnormally delayed protein

LOC100651975 -0.179626958 0.000784183 0.027064136 NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 9

232 LOC100650266 -0.176868055 0.000685679 0.024959872 actin-related protein 2

LOC100644653 -0.176700983 0.001821359 0.047657241 rapamycin-insensitive companion of mTOR

LOC100643905 -0.17536743 0.00033814 0.016268835 piezo-type mechanosensitive ion channel component

LOC100643022 -0.172646663 0.000639384 0.024076341 protein croquemort

LOC100650084 -0.170286044 0.001464491 0.040756015 uncharacterized LOC100650084

LOC100645847 -0.167405514 0.001332213 0.037840345 uncharacterized LOC100645847

LOC100644187 -0.16519472 0.001093443 0.033109325 60S ribosomal protein L36

LOC100647240 -0.164936073 0.000947646 0.030532976 succinate–CoA ligase [ADP-forming] subunit beta, mitochondrial Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC105665640 -0.161592232 0.001525584 0.041851417 NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 13

LOC100651113 -0.157767019 3.72E-06 0.000547773 growth hormone-inducible transmembrane protein

LOC100648985 -0.156641446 0.000544583 0.021668563 succinate-semialdehyde dehydrogenase [NADP(+)] GabD

LOC105665770 -0.153276005 0.001498015 0.041330612 interaptin

LOC100647329 -0.149671512 0.000586994 0.023070052 sequestosome-1

233 LOC100650471 -0.1488744 0.001156455 0.034331728 regulating synaptic membrane exocytosis protein 1

LOC100642755 -0.148625198 5.74E-05 0.00453151 N-terminal kinase-like protein

LOC100646748 -0.1476797 1.69E-05 0.001767399 GTP-binding nuclear protein Ran

LOC100644700 -0.14644207 0.001628987 0.044184562 ubiquitin-like domain-containing CTD phosphatase 1

LOC100649012 -0.145746141 0.001937971 0.049894978 small glutamine-rich tetratricopeptide repeat-containing protein alpha

LOC100649632 -0.143923557 0.000517837 0.021213876 nucleolar and coiled-body phosphoprotein 1

LOC100652257 -0.142388353 0.000696415 0.025182455 50S ribosomal protein L1

LOC100648250 -0.137985041 0.00083715 0.028483799 NADH-quinone oxidoreductase subunit B 2 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100651068 -0.133393351 0.001377041 0.03899861 60S ribosomal protein L11

LOC100648519 -0.131878189 0.000651753 0.024230605 mitochondrial folate transporter/carrier

LOC100648221 -0.131065917 0.001416573 0.039883572 serine-arginine protein 55

LOC100652182 -0.122419459 0.000131334 0.00839778 E3 ubiquitin-protein ligase RNF126-A

LOC100649628 -0.108404918 0.001092882 0.033109325 plasminogen activator inhibitor 1 RNA-binding protein

234 LOC100645346 -0.076754761 0.001547202 0.042203983 vesicle-associated membrane protein 2

LOC100642781 0.070366965 0.000426049 0.019068382 eukaryotic initiation factor 4A-I

LOC100652227 0.097893464 0.000973483 0.031038647 proline dehydrogenase 1, mitochondrial

LOC100644823 0.098448007 0.000899204 0.029652175 V-type proton ATPase subunit B

LOC100643517 0.115751188 0.001275224 0.036493644 polyubiquitin

LOC110119486 0.121967529 0.000524961 0.021314962 uncharacterized LOC110119486

LOC100647789 0.128899025 0.000906351 0.029768749 signal transducer and activator of transcription 5A

LOC100652031 0.129465385 0.000322724 0.015854657 ubiquitin-protein ligase E3B Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100647526 0.133116115 0.000771685 0.02672862 dynein light chain 2, cytoplasmic

LOC100651784 0.13335477 0.000592324 0.023091062 uncharacterized LOC100651784

LOC105667204 0.136145149 0.000744845 0.025899671 pyruvate kinase

LOC100647790 0.138645451 0.000648889 0.024217627 uncharacterized LOC100647790

LOC100650893 0.13978929 0.00070497 0.025182455 FIT family protein CG10671

235 LOC100648533 0.14066667 0.000203408 0.011521258 enolase

LOC100643282 0.146368178 0.001774878 0.046822728 acetylcholine receptor subunit beta-like 2

LOC100645942 0.150868991 0.000383202 0.0179119 double-strand-break repair protein rad21 homolog

LOC100650479 0.1551343 0.000950594 0.030532976 glutamyl aminopeptidase

LOC105666469 0.156301052 0.001539617 0.04211641 signal recognition particle subunit SRP68

LOC100644809 0.158434684 0.001854051 0.048250418 renin receptor

LOC100651650 0.159790691 0.000634381 0.024076341 sorting nexin-17

LOC100644958 0.160524375 0.000611987 0.023477384 sialin Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100650430 0.161747777 0.000174747 0.010267185 zinc ﬁnger MYM-type protein 3

LOC100648377 0.161801652 0.000110185 0.007481114 sorting nexin-14

LOC100651379 0.16217019 7.12E-05 0.005483177 glycogen debranching enzyme

LOC100647195 0.163858158 0.001192271 0.035108191 glycogen [starch] synthase

LOC100645486 0.164359615 0.001158772 0.034331728 tripartite motif-containing protein 2

236 LOC100647440 0.166292141 0.001849756 0.048250418 nuclear hormone receptor HR96

LOC100648480 0.168576512 0.001223581 0.03581113 solute carrier organic anion transporter family member 4A1

LOC100643214 0.169854709 0.0019608 0.049980775 4-coumarate–CoA ligase 1

LOC100650874 0.170284633 0.001088561 0.033109325 uncharacterized LOC100650874

LOC100643424 0.170566876 0.001228755 0.035853589 RWD domain-containing protein 1

LOC100642289 0.174211853 0.000159373 0.009591261 equilibrative nucleoside transporter 4

LOC100642807 0.174514384 3.45E-05 0.003043379 icarapin-like

LOC100650043 0.178936914 0.000388236 0.017935807 calcium-activated potassium channel slowpoke Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100646598 0.183353946 0.00080659 0.027738056 pyrroline-5-carboxylate reductase 2

LOC100647296 0.183477018 5.92E-09 2.53E-06 translation elongation factor 2

LOC100651332 0.183835782 3.95E-08 1.27E-05 cAMP-dependent protein kinase type I regulatory subunit

LOC100644848 0.185055698 1.12E-05 0.001283457 ATP-binding cassette sub-family G member 1

LOC100651600 0.185612229 4.46E-06 0.000597014 probable maleylacetoacetate isomerase 2

237 LOC100648328 0.186022679 4.00E-06 0.000555066 juvenile hormone epoxide hydrolase 1

LOC100650393 0.187947461 0.000531269 0.021314962 cyclic AMP-dependent transcription factor ATF-6 alpha

LOC100642954 0.188193325 1.51E-05 0.001636395 proton-coupled amino acid transporter-like protein CG1139

LOC100649204 0.189788318 0.000150974 0.009318793 uncharacterized LOC100649204

LOC100650098 0.190717679 0.001796248 0.047128273 constitutive coactivator of PPAR-gamma-like protein 1 homolog

LOC100648873 0.193574166 0.001274948 0.036493644 rab GTPase-binding effector protein 1

LOC100643772 0.19374974 0.001452211 0.040611623 uncharacterized LOC100643772

LOC100649938 0.194632905 0.00189648 0.048957656 thymosin beta Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100645881 0.197639305 0.000861943 0.029123604 very-long-chain (3R)-3-hydroxyacyl-CoA dehydratase hpo-8

LOC100650366 0.197936315 0.001702251 0.045913089 xanthine dehydrogenase

LOC100647029 0.200163387 0.001010398 0.031501354 protein kinase C-binding protein 1

LOC100650975 0.201416517 4.47E-05 0.003675708 methionine aminopeptidase 2

LOC100650166 0.209864839 1.24E-05 0.001401238 1,4-alpha-glucan-branching enzyme

238 LOC100651862 0.21080825 0.0010126 0.031501354 3-ketoacyl-CoA thiolase, mitochondrial

LOC100644444 0.212515927 6.86E-06 0.000858465 pregnancy zone protein

LOC100651873 0.213221796 0.001957854 0.049980775 RNA-binding protein Rsf1

LOC100649035 0.214108629 3.11E-05 0.002856481 BCL2/adenovirus E1B 19 kDa protein-interacting protein 3

LOC100650064 0.214167631 7.10E-07 0.000133976 mitogen-activated protein kinase kinase kinase 4

LOC100647245 0.214191745 2.70E-07 6.18E-05 mucolipin-3

LOC100642398 0.214847298 0.000741549 0.025899671 dynein regulatory complex protein 1

LOC100650814 0.214974812 0.000125936 0.008138501 uncharacterized LOC100650814 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100646258 0.217949476 0.000180971 0.010546494 USP6 N-terminal-like protein

LOC100646337 0.221748332 0.000270701 0.013791423 CBP80/20-dependent translation initiation factor

LOC100648357 0.223165462 0.000706123 0.025182455 kxDL motif-containing protein CG10681

LOC105665615 0.224000268 0.000997874 0.031501354 DNA repair protein REV1

LOC100644465 0.224509092 7.25E-05 0.005537263 dynein assembly factor 5, axonemal

239 LOC100642744 0.225899888 4.55E-05 0.00371624 45 kDa calcium-binding protein

LOC100642912 0.226290775 9.87E-05 0.006988246 probable phospholipid hydroperoxide glutathione peroxidase

LOC100649904 0.229046147 1.53E-07 3.69E-05 CD63 antigen

LOC100642906 0.231722975 0.001083245 0.033109325 aminopeptidase Ey

LOC100643245 0.232455417 0.001385058 0.039110617 RAB6A-GEF complex partner protein 2

LOC100651936 0.232902509 7.50E-05 0.005640796 cytoplasmic aconitate hydratase

LOC100649947 0.234117085 4.89E-05 0.003953257 short/branched chain speciﬁc acyl-CoA dehydrogenase, mitochondria

LOC100645683 0.234793249 0.000222522 0.012224224 UDP-glucuronosyltransferase 2B30 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100642613 0.236280649 0.000666489 0.024588603 uncharacterized LOC100642613

LOC100643252 0.240155093 0.000877807 0.02949671 phospholipid scramblase 2

LOC100646374 0.244599856 0.00099279 0.031501354 battenin

LOC100646172 0.245743689 0.000325387 0.015904328 condensin-2 complex subunit D3

LOC100646402 0.246752289 8.53E-05 0.006317176 voltage-gated potassium channel subunit beta-2

240 LOC100647255 0.247977373 5.40E-07 0.00011065 guanine nucleotide-binding protein subunit beta-like protein

LOC100651168 0.254177603 0.001709272 0.045973689 heterogeneous nuclear ribonucleoprotein A3 homolog 2

LOC100647089 0.255398484 0.001521816 0.041851417 cell division cycle protein 23 homolog

LOC100644921 0.257490453 7.94E-06 0.000979693 proton-coupled amino acid transporter-like protein CG1139

LOC100650324 0.260631492 0.00044978 0.01942121 glycine cleavage system H protein, mitochondrial

LOC100648603 0.26319238 9.00E-09 3.61E-06 uncharacterized phosphotransferase YvkC

LOC105666388 0.263555555 5.09E-06 0.000671083 adenosine deaminase CECR1

LOC100642454 0.26391567 0.000389301 0.017935807 L-xylulose reductase Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100643441 0.270383369 3.83E-05 0.003262415 lipoyltransferase 1, mitochondrial

LOC100650568 0.273584359 0.001239782 0.036066057 open rectiﬁer potassium channel protein 1

LOC105666948 0.274193969 0.001770357 0.046822728 sodium-coupled monocarboxylate transporter 1

LOC100650108 0.275238249 1.10E-05 0.001283457 multidrug resistance protein homolog 49

LOC100645099 0.276556375 0.000428717 0.019068382 origin recognition complex subunit 4

241 LOC100651089 0.277393504 6.78E-05 0.005268326 uncharacterized LOC100651089

LOC100648833 0.278143639 3.56E-05 0.003114585 alaserpin

LOC100645243 0.279151679 6.71E-08 1.96E-05 putative 28S ribosomal protein S5, mitochondrial

LOC100645499 0.281700538 0.000359669 0.01697675 palmitoyl-protein thioesterase 1

LOC100642931 0.282359814 8.62E-07 0.000156628 glycogen phosphorylase

LOC100651992 0.284057381 2.18E-08 7.77E-06 FACT complex subunit Ssrp1

LOC105665917 0.284141904 8.79E-05 0.006463696 uncharacterized LOC105665917

LOC105666557 0.285353654 3.75E-06 0.000547773 heparanase Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100652086 0.28607429 2.76E-05 0.002628623 probable RNA-binding protein 19

LOC100644950 0.292042399 3.81E-05 0.003262415 CXXC-type zinc ﬁnger protein 1

LOC110119297 0.303032131 0.001474908 0.04092763 uncharacterized LOC110119297

LOC105666068 0.303859525 0.000141198 0.008849658 uncharacterized LOC105666068

LOC100650995 0.306769571 0.000202425 0.011521258 solute carrier family 23 member 1

242 LOC100642900 0.3077201 0.000646294 0.024214665 biogenesis of lysosome-related organelles complex 1 subunit 6

LOC100652007 0.323601939 4.37E-05 0.003623571 carbohydrate sulfotransferase 11

LOC100646260 0.324822239 3.27E-07 6.85E-05 receptor expression-enhancing protein 5

LOC100644229 0.325476167 5.69E-09 2.53E-06 heat shock 70 kDa protein cognate 4

LOC100644112 0.328813879 0.000951282 0.030532976 fructose-bisphosphate aldolase

LOC100645868 0.331730774 0.000234415 0.012591147 putative leucine-rich repeat-containing protein DDB_G0290503

LOC100648599 0.331894929 0.000731612 0.025710557 proton-coupled amino acid transporter-like protein CG1139

LOC100642680 0.337029953 1.64E-05 0.001739897 leucine-rich repeat-containing protein 20 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100651052 0.343183694 0.00021168 0.011850383 glyoxylate reductase/hydroxypyruvate reductase

LOC100647856 0.346645259 0.000224191 0.012224224 uncharacterized LOC100647856

LOC100645813 0.351310366 6.78E-07 0.000130499 dentin sialophosphoprotein

LOC100650392 0.353623918 0.000286787 0.01445798 ATP-binding cassette sub-family A member 13

LOC100642530 0.356712238 0.000429769 0.019068382 sodium-independent sulfate anion transporter

243 LOC100645349 0.360426438 0.000248809 0.013091708 spermine oxidase

LOC100646865 0.361037257 0.000394077 0.018069381 calcium release-activated calcium channel protein 1

LOC100647502 0.36566943 0.001115036 0.033447612 uncharacterized LOC100647502

LOC100645209 0.367267838 1.44E-10 8.65E-08 multidrug resistance-associated protein 4

LOC100650119 0.367696948 4.04E-06 0.000555066 protein cueball

LOC100646721 0.368704481 0.000488715 0.020549502 venom acid phosphatase Acph-1

LOC105666521 0.370158593 0.000200498 0.011491648 uncharacterized LOC105666521

LOC100650431 0.371323892 3.15E-07 6.83E-05 jmjC domain-containing histone demethylation protein 1 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC105666757 0.372249817 0.000718674 0.025535467 uncharacterized LOC105666757

LOC100645679 0.373864837 2.32E-05 0.00223759 peptide transporter family 1

LOC100645828 0.375570223 3.40E-06 0.000511369 dystroglycan

LOC100643283 0.380925017 0.000224705 0.012224224 pyruvate dehydrogenase E1 component subunit beta, mitochondrial

LOC100650362 0.384271698 0.000528989 0.021314962 uncharacterized LOC100650362

244 LOC100644743 0.388223846 0.000435668 0.019068382 cytochrome P450 4c3

LOC100645241 0.393751393 0.001273009 0.036493644 estradiol 17-beta-dehydrogenase 2

LOC100649178 0.395556559 1.31E-05 0.001439564 ras-related and estrogen-regulated growth inhibitor

LOC100643580 0.397141668 4.26E-05 0.00357097 purine nucleoside phosphorylase

LOC100650900 0.411899268 0.000640102 0.024076341 dynactin, 150 kDa isoform

LOC100647796 0.42299568 1.27E-05 0.001423863 uncharacterized LOC100647796

LOC100648867 0.430184864 0.000569577 0.022477282 uncharacterized LOC100648867

LOC100651847 0.434877883 0.000828976 0.028305725 cytosolic carboxypeptidase 2 Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100650307 0.436344032 2.28E-05 0.00222001 uncharacterized LOC100650307

LOC100644732 0.437817416 2.72E-17 6.56E-14 protein eiger

LOC100650287 0.440495518 3.11E-05 0.002856481 replication protein A 32 kDa subunit

LOC100647179 0.45977522 8.05E-07 0.000149155 uncharacterized LOC100647179

LOC100650345 0.474553828 0.001079688 0.033109271 ejaculatory bulb-speciﬁc protein 3

245 LOC105665782 0.480328515 0.000417583 0.018966524 uncharacterized LOC105665782

LOC100650153 0.495859834 3.00E-08 9.95E-06 protein inscuteable homolog

LOC100645152 0.499041434 0.000222523 0.012224224 nucleolar protein 14-like

LOC100649414 0.500212746 1.57E-08 5.81E-06 Bardet-Biedl syndrome 2 protein homolog

LOC105666310 0.509839557 0.000467765 0.019929676 uncharacterized LOC105666310

LOC100646382 0.524598081 2.84E-05 0.002650766 ras-related and estrogen-regulated growth inhibitor

LOC105666115 0.526157786 0.000499897 0.020837716 uncharacterized LOC105666115

LOC100649714 0.561957001 0.000350983 0.016648373 calcyphosin-like protein Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100645386 0.562403575 7.69E-12 7.24E-09 protein 5NUC

LOC100648955 0.565849567 0.000596125 0.023145532 probable peroxisomal acyl-coenzyme A oxidase 1

LOC100647051 0.576341242 0.001961444 0.049980775 beta-ureidopropionase

LOC100650650 0.5812901 2.68E-21 1.29E-17 cartilage oligomeric matrix protein

LOC100651799 0.590982123 1.58E-06 0.000272341 uncharacterized LOC100651799

246 LOC100648691 0.619239668 6.05E-09 2.53E-06 sodium-dependent nutrient amino acid transporter 1

LOC100650597 0.780208084 3.34E-10 1.89E-07 pyrokinin-1 receptor

LOC100650704 0.785148399 6.12E-07 0.000122693 fructose-1,6-bisphosphatase 1

LOC100642739 0.859441994 4.11E-10 2.20E-07 facilitated trehalose transporter Tret1-2 homolog

LOC105666635 0.890312967 5.81E-05 0.004546492 uncharacterized LOC105666635

LOC100648391 0.908856792 8.74E-26 8.42E-22 cytochrome P450 6k1

LOC100631088 0.949388981 5.98E-09 2.53E-06 phosphoenolpyruvate carboxykinase

LOC100645024 1.015416098 4.39E-16 8.45E-13 apyrase Table A.3 continued from previous page

LOC log2FoldChange pvalue padj name

LOC100646400 1.100608093 3.92E-11 2.90E-08 ionotropic receptor 25a

LOC105666404 1.104916952 6.57E-06 0.00083252 troponin C

LOC100642296 1.1424906 1.13E-07 2.93E-05 keratin-associated protein 19-2 247 TABLE A.4: These tables show gene ontology terms and biological processes for genes that differentially expressed between treated and untreated bees.

Gene ontology term P-adj Biological processes

GO:0055114 4.16E-05 oxidation-reduction process

GO:0070873 0.0003589749 regulation of glycogen metabolic process

GO:1900074 0.0005143576 negative regulation of neuromuscular synaptic transmission

GO:0006073 0.0006517591 cellular glucan metabolic process

GO:0006112 0.0006517591 energy reserve metabolic process 248 GO:0098656 0.0006915762 anion transmembrane transport

GO:0003333 0.0009199461 amino acid transmembrane transport

GO:0046942 0.0011163271 carboxylic acid transport

GO:0006091 0.0013792863 generation of precursor metabolites and energy

GO:0019362 0.0013808232 pyridine nucleotide metabolic process

GO:1903825 0.0014037256 organic acid transmembrane transport

GO:0072376 0.0014414694 protein activation cascade

GO:1902109 0.0014414694 negative regulation of mitochondrial membrane permeability involved in apoptotic process Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0006540 0.0014414694 glutamate decarboxylation to succinate

GO:0006681 0.0014414694 galactosylceramide metabolic process

GO:0045819 0.0014414694 positive regulation of glycogen catabolic process

GO:1905709 0.0014414694 negative regulation of membrane permeability

GO:0030449 0.0014414694 regulation of complement activation

249 GO:0005978 0.0015163303 glycogen biosynthetic process

GO:0001938 0.0015163303 positive regulation of endothelial cell proliferation

GO:0055085 0.001609144 transmembrane transport

GO:0016051 0.0016844941 carbohydrate biosynthetic process

GO:0010907 0.0017004361 positive regulation of glucose metabolic process

GO:0043471 0.0020063814 regulation of cellular carbohydrate catabolic process

GO:1901564 0.0023091026 organonitrogen compound metabolic process

GO:0035966 0.0024984422 response to topologically incorrect protein Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0048132 0.0026444132 female germ-line stem cell asymmetric division

GO:0044247 0.0026444132 cellular polysaccharide catabolic process

GO:0002526 0.0026444132 acute inﬂammatory response

GO:0009251 0.0026444132 glucan catabolic process

GO:0044712 0.003224437 single-organism catabolic process

250 GO:0045579 0.0041890472 positive regulation of B cell differentiation

GO:0001178 0.0042154747 regulation of transcriptional start site selection at RNA polymerase II promoter

positive regulation of transcription from RNA polymerase II promoter involved in unfolded GO:0006990 0.0042154747 protein response

GO:1903715 0.0042154747 regulation of aerobic respiration

GO:0070327 0.0042154747 thyroid hormone transport

GO:0017121 0.0042154747 phospholipid scrambling

GO:0070328 0.0053065797 triglyceride homeostasis Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0007566 0.0053541726 embryo implantation

GO:0016052 0.005602937 carbohydrate catabolic process

GO:1902107 0.005794437 positive regulation of leukocyte differentiation

GO:0046031 0.0060066213 ADP metabolic process

GO:0033692 0.0061297209 cellular polysaccharide biosynthetic process

251 GO:0034620 0.0066583019 cellular response to unfolded protein

GO:0055088 0.0067147591 lipid homeostasis

GO:0045912 0.0071567561 negative regulation of carbohydrate metabolic process

GO:0002181 0.0073578892 cytoplasmic translation

GO:0002696 0.0074815508 positive regulation of leukocyte activation

GO:0009185 0.0074815508 ribonucleoside diphosphate metabolic process

GO:0009135 0.0074815508 purine nucleoside diphosphate metabolic process

GO:0043603 0.0076364717 cellular amide metabolic process Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0001173 0.0082192367 DNA-templated transcriptional start site selection

GO:0045820 0.0082192367 negative regulation of glycolytic process

GO:0051195 0.0082192367 negative regulation of cofactor metabolic process

GO:0007042 0.0082192367 lysosomal lumen acidiﬁcation

GO:0044699 0.0084251459 single-organism process

252 GO:0005976 0.0092014835 polysaccharide metabolic process

GO:0006165 0.0092014835 nucleoside diphosphate phosphorylation

GO:0006536 0.0102439676 glutamate metabolic process

GO:0045619 0.0119324595 regulation of lymphocyte differentiation

GO:0065002 0.0123531941 intracellular protein transmembrane transport

GO:0060055 0.0133558227 angiogenesis involved in wound healing

GO:0009065 0.0145092184 glutamine family amino acid catabolic process

GO:0030150 0.0145092184 protein import into mitochondrial matrix Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0009267 0.0151583814 cellular response to starvation

GO:0042632 0.0158204056 cholesterol homeostasis

GO:0046034 0.0166453135 ATP metabolic process

GO:0015672 0.0169206208 monovalent inorganic cation transport

GO:0050727 0.0170067616 regulation of inﬂammatory response

253 GO:0019752 0.0171022216 carboxylic acid metabolic process

GO:0019233 0.0176349276 sensory perception of pain

GO:0051897 0.0176349276 positive regulation of protein kinase B signaling

GO:0050865 0.0191096568 regulation of cell activation

GO:0048025 0.0195338527 negative regulation of mRNA splicing, via spliceosome

GO:0006678 0.0195338527 glucosylceramide metabolic process

GO:0035456 0.0195338527 response to interferon-beta

GO:0042491 0.0195338527 auditory receptor cell differentiation Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0050870 0.0204216623 positive regulation of T cell activation

GO:0045913 0.0211054689 positive regulation of carbohydrate metabolic process

GO:0042078 0.0211054689 germ-line stem cell division

GO:0016054 0.021460426 organic acid catabolic process

GO:0008219 0.0223989831 cell death

254 GO:0051289 0.0229997292 protein homotetramerization

GO:0019318 0.0241988254 hexose metabolic process

GO:0032781 0.0249221665 positive regulation of ATPase activity

GO:0061024 0.0252762212 membrane organization

GO:0044765 0.0253732789 single-organism transport

GO:0009063 0.0257525518 cellular amino acid catabolic process

GO:0006749 0.0257669664 glutathione metabolic process

GO:0097345 0.0266672194 mitochondrial outer membrane permeabilization Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0045071 0.0266672194 negative regulation of viral genome replication

GO:1902686 0.0266672194 mitochondrial outer membrane permeabilization involved in programmed cell death

GO:0006560 0.0266672194 proline metabolic process

GO:0098700 0.0266672194 neurotransmitter loading into synaptic vesicle

GO:1905710 0.0266672194 positive regulation of membrane permeability

255 GO:0072593 0.0276838771 reactive oxygen species metabolic process

GO:0034112 0.0287258596 positive regulation of homotypic cell-cell adhesion

GO:0015804 0.0290845478 neutral amino acid transport

GO:0050864 0.0290845478 regulation of B cell activation

GO:0042059 0.0290845478 negative regulation of epidermal growth factor receptor signaling pathway

GO:0016485 0.0312516403 protein processing

GO:0006915 0.031484691 apoptotic process

GO:0008637 0.0318783679 apoptotic mitochondrial changes Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0043043 0.0340556004 peptide biosynthetic process

GO:0006820 0.0343062565 anion transport

GO:0034976 0.0343325415 response to endoplasmic reticulum stress

GO:0035924 0.0346748215 cellular response to vascular endothelial growth factor stimulus

GO:0071248 0.0352259354 cellular response to metal ion

256 GO:0031668 0.0361391764 cellular response to extracellular stimulus

GO:0048102 0.0377500944 autophagic cell death

GO:0045333 0.0380089586 cellular respiration

GO:2001261 0.0380271084 negative regulation of semaphorin-plexin signaling pathway

GO:0043103 0.0380271084 hypoxanthine salvage

GO:2000825 0.0380271084 positive regulation of androgen receptor activity

GO:2000510 0.0380271084 positive regulation of dendritic cell chemotaxis

GO:0072684 0.0380271084 mitochondrial tRNA 3’-trailer cleavage, endonucleolytic Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0048382 0.0380271084 mesendoderm development

GO:0019100 0.0380271084 male germ-line sex determination

GO:0032803 0.0380271084 regulation of low-density lipoprotein particle receptor catabolic process

GO:0060718 0.0380271084 chorionic trophoblast cell differentiation

GO:0032695 0.0380271084 negative regulation of interleukin-12 production

257 GO:0032689 0.0380271084 negative regulation of interferon-gamma production

GO:0045052 0.0380271084 protein insertion into ER membrane by GPI attachment sequence

GO:0060691 0.0380271084 epithelial cell maturation involved in salivary gland development

GO:0097254 0.0380271084 renal tubular secretion

GO:0060689 0.0380271084 cell differentiation involved in salivary gland development

GO:0006422 0.0380271084 aspartyl-tRNA aminoacylation

GO:0006114 0.0380271084 glycerol biosynthetic process

GO:0045348 0.0380271084 positive regulation of MHC class II biosynthetic process Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0048601 0.0380271084 oocyte morphogenesis

GO:1902952 0.0380271084 positive regulation of dendritic spine maintenance

GO:0045163 0.0380271084 clustering of voltage-gated potassium channels

GO:0002005 0.0380271084 angiotensin catabolic process in blood

GO:0051078 0.0380271084 meiotic nuclear envelope disassembly

258 GO:0006600 0.0380271084 creatine metabolic process

GO:0044130 0.0380271084 negative regulation of growth of symbiont in host

GO:0045852 0.0380271084 pH elevation

GO:0044266 0.0380271084 multicellular organismal macromolecule catabolic process

GO:0044268 0.0380271084 multicellular organismal protein metabolic process

GO:0044256 0.0380271084 protein digestion

GO:0045586 0.0380271084 regulation of gamma-delta T cell differentiation

GO:0015801 0.0380271084 aromatic amino acid transport Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0006958 0.0380271084 complement activation, classical pathway

GO:0045647 0.0380271084 negative regulation of erythrocyte differentiation

GO:1903489 0.0380271084 positive regulation of lactation

GO:1903441 0.0380271084 protein localization to ciliary membrane

GO:0002206 0.0380271084 gene conversion of immunoglobulin genes

259 GO:1903758 0.0380271084 negative regulation of transcription from RNA polymerase II promoter by histone modiﬁcation

GO:0007384 0.0380271084 speciﬁcation of segmental identity, thorax

GO:0038161 0.0380271084 prolactin signaling pathway

GO:0007429 0.0380271084 secondary branching, open tracheal system

GO:1990418 0.0380271084 response to insulin-like growth factor stimulus

GO:1990180 0.0380271084 mitochondrial tRNA 3’-end processing

GO:1990146 0.0380271084 protein localization to rhabdomere

GO:0035356 0.0380271084 cellular triglyceride homeostasis Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0035470 0.0380271084 positive regulation of vascular wound healing

GO:0016243 0.0380271084 regulation of autophagosome size

GO:0034356 0.0380271084 NAD biosynthesis via nicotinamide riboside salvage pathway

GO:0046595 0.0380271084 establishment of pole plasm mRNA localization

GO:0035229 0.0380271084 positive regulation of glutamate-cysteine ligase activity

260 GO:0010989 0.0380271084 negative regulation of low-density lipoprotein particle clearance

GO:0046544 0.0380271084 development of secondary male sexual characteristics

GO:0035752 0.0380271084 lysosomal lumen pH elevation

GO:0046629 0.0380271084 gamma-delta T cell activation

GO:1905267 0.0380271084 endonucleolytic cleavage involved in tRNA processing

GO:0034721 0.0380271084 histone H3-K4 demethylation, trimethyl-H3-K4-speciﬁc

GO:1905719 0.0380271084 protein localization to perinuclear region of cytoplasm

GO:0071499 0.0380271084 cellular response to laminar ﬂuid shear stress Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0035694 0.0380271084 mitochondrial protein catabolic process

GO:1905802 0.0380271084 regulation of cellular response to manganese ion

GO:0042104 0.0380271084 positive regulation of activated T cell proliferation

GO:0070481 0.0380271084 nuclear-transcribed mRNA catabolic process, non-stop decay

GO:0071205 0.0380271084 protein localization to juxtaparanode region of axon

261 GO:0009115 0.0380271084 xanthine catabolic process

GO:0031426 0.0380271084 polycistronic mRNA processing

GO:0042779 0.0380271084 tRNA 3’-trailer cleavage

GO:0009245 0.0380271084 lipid A biosynthetic process

GO:0071579 0.0380271084 regulation of zinc ion transport

GO:0071629 0.0380271084 ubiquitin-dependent catabolism of misfolded proteins by cytoplasm-associated proteasome

GO:1901165 0.0380271084 positive regulation of trophoblast cell migration

GO:0070837 0.0380271084 dehydroascorbic acid transport Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:1901269 0.0380271084 lipooligosaccharide metabolic process

GO:0070651 0.0380271084 nonfunctional rRNA decay

GO:1900100 0.0380271084 positive regulation of plasma cell differentiation

GO:0036324 0.0380271084 vascular endothelial growth factor receptor-2 signaling pathway

GO:0042908 0.0380271084 xenobiotic transport

262 GO:0070904 0.0380271084 transepithelial L-ascorbic acid transport

GO:2000347 0.0380271084 positive regulation of hepatocyte proliferation

GO:0044281 0.0381769803 small molecule metabolic process

GO:0009144 0.0404458409 purine nucleoside triphosphate metabolic process

GO:0009896 0.0404458409 positive regulation of catabolic process

GO:0070887 0.040614941 cellular response to chemical stimulus

GO:0051179 0.0408983636 localization

GO:0009199 0.0421339443 ribonucleoside triphosphate metabolic process Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0009126 0.0421339443 purine nucleoside monophosphate metabolic process

GO:0043467 0.0423233578 regulation of generation of precursor metabolites and energy

GO:0071496 0.0423717475 cellular response to external stimulus

GO:0007006 0.0425095354 mitochondrial membrane organization

GO:0032402 0.0434803095 melanosome transport

263 GO:0043649 0.0434803095 dicarboxylic acid catabolic process

GO:0006111 0.0434803095 regulation of gluconeogenesis

GO:0007638 0.0434803095 mechanosensory behavior

GO:0042987 0.0434803095 amyloid precursor protein catabolic process

GO:0060326 0.0436177777 cell chemotaxis

GO:0048525 0.0436177777 negative regulation of viral process

GO:0035071 0.0436177777 salivary gland cell autophagic cell death

GO:0019058 0.0438928481 viral life cycle Table A.4 continued from previous page

Gene ontology term P-adj Biological processes

GO:0006096 0.0455832456 glycolytic process

GO:0021700 0.047880496 developmental maturation 264 TABLE A.5: These tables show gene ID and description for genes that deferentially expressed transcripts between treated and untreated bees

Loc q-value Gene Name

LOC100650650 0.002919439 cartilage oligomeric matrix protein

LOC100651433 0.00582838 protein lethal(2)essential for life

LOC100648391 0.00582838 cytochrome P450 6k1

LOC100644229 0.00582838 heat shock 70 kDa protein cognate 4

LOC100647296 0.006588435 translation elongation factor 2 265

LOC110119287 0.00791348 neutral and basic amino acid transport protein rBAT

LOC100646534 0.009801357 L-asparaginase

LOC100643972 0.009801357 proton-coupled amino acid transporter-like protein pathetic

LOC100642770 0.009801357 eukaryotic translation initiation factor 2 subunit 2

LOC100642739 0.009801357 facilitated trehalose transporter Tret1-2 homolog

LOC100647245 0.009801357 mucolipin-3

LOC100646094 0.027580135 peptide methionine sulfoxide reductase Table A.5 continued from previous page

Loc q-value Gene Name

LOC100648603 0.027992828 phosphotransferase YvkC

LOC100649904 0.027992828 CD63 antigen

LOC100644848 0.030726569 ATP-binding cassette sub-family G member 1

LOC100645243 0.035245918 putative 28S ribosomal protein S5, mitochondrial

LOC100650338 0.039612435 transcription factor CP2-like protein 1

266 LOC100648549 0.043982678 cytochrome c

LOC100649106 0.043982678 putative glutathione-speciﬁc gamma-glutamylcyclotransferase 2

LOC100650975 0.043982678 methionine aminopeptidase 2

LOC100645983 0.044400748 large neutral amino acids transporter small subunit 2

LOC100646748 0.044400748 GTP-binding nuclear protein Ran

LOC100643135 0.044400748 protein l(2)37Cc

LOC100650710 0.044400748 general odorant-binding protein 71

LOC100642807 0.047010973 icarapin-like TABLE A.6: These tables show GO terms and biological processes for genes that differentially expressed transcripts between treated and untreated bees.

Gene ontology terms P-value Biological processes

GO:0008514 0.000218668 organic anion transmembrane transporter activity

GO:0051082 0.000630391 unfolded protein binding

GO:0005342 0.001367436 organic acid transmembrane transporter activity

GO:0004775 0.001978135 succinate-CoA ligase (ADP-forming) activity

GO:0016491 0.004670896 oxidoreductase activity 267 GO:0051864 0.006370351 histone demethylase activity (H3-K36 speciﬁc)

GO:0015171 0.007083489 amino acid transmembrane transporter activity

GO:0051287 0.00710752 NAD binding

GO:0048037 0.007506738 cofactor binding

GO:0015291 0.010509612 secondary active transmembrane transporter activity

GO:0016209 0.01491312 antioxidant activity

GO:0004563 0.016942749 beta-N-acetylhexosaminidase activity

GO:0016758 0.017001143 transferase activity, transferring hexosyl groups Table A.6 continued from previous page

Gene ontology terms P-value Biological processes

GO:0003680 0.021414831 AT DNA binding

GO:0016405 0.021414831 CoA-ligase activity

GO:0019534 0.021414831 toxin transporter activity

GO:0016655 0.02336168 oxidoreductase activity, acting on NAD(P)H, quinone or similar compound as acceptor

GO:0008454 0.02596401 alpha-1,3-mannosylglycoprotein 4-beta-N-acetylglucosaminyltransferase activity

268 GO:0008430 0.02596401 selenium binding

GO:0008184 0.02596401 glycogen phosphorylase activity

GO:0008113 0.02596401 peptide-methionine (S)-S-oxide reductase activity

GO:0000035 0.02596401 acyl binding

GO:0000036 0.02596401 ACP phosphopantetheine attachment site binding involved in fatty acid biosynthetic process

GO:0047724 0.02596401 inosine nucleosidase activity

GO:0042132 0.02596401 fructose 1,6-bisphosphate 1-phosphatase activity

GO:0030984 0.02596401 kininogen binding Table A.6 continued from previous page

Gene ontology terms P-value Biological processes

GO:0008747 0.02596401 N-acetylneuraminate lyase activity

GO:0043136 0.02596401 glycerol-3-phosphatase activity

GO:0035226 0.02596401 glutamate-cysteine ligase catalytic subunit binding

GO:0004157 0.02596401 dihydropyrimidinase activity

GO:0050221 0.02596401 prostaglandin-E2 9-reductase activity

269 GO:0008559 0.02596401 xenobiotic-transporting ATPase activity

GO:0008520 0.02596401 L-ascorbate:sodium symporter activity

GO:0008517 0.02596401 folic acid transporter activity

GO:0004777 0.02596401 succinate-semialdehyde dehydrogenase (NAD+) activity

GO:0004778 0.02596401 succinyl-CoA hydrolase activity

GO:0004740 0.02596401 pyruvate dehydrogenase (acetyl-transferring) kinase activity

GO:0003844 0.02596401 1,4-alpha-glucan branching enzyme activity

GO:0009013 0.02596401 succinate-semialdehyde dehydrogenase [NAD(P)+] activity Table A.6 continued from previous page

Gene ontology terms P-value Biological processes

GO:0000121 0.02596401 glycerol-1-phosphatase activity

GO:0004581 0.02596401 dolichyl-phosphate beta-glucosyltransferase activity

GO:0004584 0.02596401 dolichyl-phosphate-mannose-glycolipid alpha-mannosyltransferase activity

GO:0004517 0.02596401 nitric-oxide synthase activity

GO:0030350 0.02596401 iron-responsive element binding

270 GO:0008967 0.02596401 phosphoglycolate phosphatase activity

GO:0004613 0.02596401 phosphoenolpyruvate carboxykinase (GTP) activity

GO:0004634 0.02596401 phosphopyruvate hydratase activity

GO:0000823 0.02596401 inositol-1,4,5-trisphosphate 6-kinase activity

GO:0000824 0.02596401 inositol tetrakisphosphate 3-kinase activity

GO:0000825 0.02596401 inositol tetrakisphosphate 6-kinase activity

GO:0031682 0.02596401 G-protein gamma-subunit binding

GO:0031685 0.02596401 adenosine receptor binding Table A.6 continued from previous page

Gene ontology terms P-value Biological processes

GO:0047021 0.02596401 15-hydroxyprostaglandin dehydrogenase (NADP+) activity

GO:0002054 0.02596401 nucleobase binding

GO:0002058 0.02596401 uracil binding

GO:0002059 0.02596401 thymine binding

GO:0051192 0.02596401 prosthetic group binding

271 GO:0070890 0.02596401 sodium-dependent L-ascorbate transmembrane transporter activity

GO:0047326 0.02596401 inositol tetrakisphosphate 5-kinase activity

GO:0050038 0.02596401 L-xylulose reductase (NADP+) activity

GO:0033961 0.02596401 cis-stilbene-oxide hydrolase activity

GO:0003723 0.029530305 RNA binding

GO:0005337 0.031622484 nucleoside transmembrane transporter activity

GO:0046982 0.032727257 protein heterodimerization activity

GO:1901677 0.035070819 phosphate transmembrane transporter activity Table A.6 continued from previous page

Gene ontology terms P-value Biological processes

GO:0000287 0.035118139 magnesium ion binding

GO:0016831 0.041864993 carboxy-lyase activity

GO:0032451 0.043352503 demethylase activity

GO:0016878 0.043352503 acid-thiol ligase activity

GO:0015075 0.044230596 ion transmembrane transporter activity

272 GO:0022892 0.04754955 substrate-speciﬁc transporter activity

GO:0020037 0.049126871 heme binding

GO:0004879 0.049731171 RNA polymerase II transcription factor activity, ligand-activated sequence-speciﬁc DNA binding

GO:0003735 0.049991938 structural constituent of ribosome Bibliography

[1] Waldan K. Kwong and Nancy A. Moran. Gut microbial communities of social bees. Nature Reviews Microbiology, 14(6):374–384, June 2016. ISSN 1740-1534. doi: 10.1038/nrmicro.2016.43. URL https://www.nature.com/articles/ nrmicro.2016.43.

[2] Daniel H. Janzen. Coevolution of Mutualism Between Ants and Acacias in Cen- tral America. Evolution, 20(3):249–275, 1966. ISSN 1558-5646. doi: 10.1111/ j.1558-5646.1966.tb03364.x. URL https://onlinelibrary.wiley.com/doi/ abs/10.1111/j.1558-5646.1966.tb03364.x.

[3] Paul R. Ehrlich and Peter H. Raven. Butterﬂies and Plants: A Study in Co- evolution. Evolution, 18(4):586–608, 1965. ISSN 1558-5646. doi: 10.1111/ j.1558-5646.1964.tb01674.x. URL https://onlinelibrary.wiley.com/doi/ abs/10.1111/j.1558-5646.1964.tb01674.x.

[4] Jens M. Olesen, Jordi Bascompte, Yoko L. Dupont, and Pedro Jordano. The modu- larity of pollination networks. Proceedings of the National Academy of Sciences of the United States of America, 104(50):19891–19896, December 2007. ISSN 1091- 6490. doi: 10.1073/pnas.0706375104.

[5] Alexandra-Maria Klein, Bernard E. Vaissière, James H. Cane, Ingolf Steffan- Dewenter, Saul A. Cunningham, Claire Kremen, and Teja Tscharntke. Impor- tance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society of London B: Biological Sciences, 274(1608):303–313, February

273 2007. ISSN 0962-8452, 1471-2954. doi: 10.1098/rspb.2006.3721. URL http: //rspb.royalsocietypublishing.org/content/274/1608/303.

[6] C. B. Williams. The Field of Research in Preventive Entomology. Annals of Applied Biology, 34(2):175–185, 1947. ISSN 1744-7348. doi: 10.1111/j.1744-7348.1947. tb06353.x. URL https://onlinelibrary.wiley.com/doi/abs/10.1111/j. 1744-7348.1947.tb06353.x.

[7] R. F.Chapman. The Insects: Structure and Function. Cambridge University Press, 2013. ISBN 9780521113892. Google-Books-ID: NXJEi8fo7CkC.

[8] WHO Scheme. Pesticides and their application : for the control of vectors and pests of public health importance. 2006. URL http://apps.who.int/iris/ handle/10665/69223.

[9] Clevo Wilson and Clem Tisdell. Why farmers continue to use pesticides despite environmental, health and sustainability costs. Ecological Economics, 39 (3):449–462, December 2001. ISSN 0921-8009. doi: 10.1016/S0921-8009(01) 00238-5. URL http://www.sciencedirect.com/science/article/pii/ S0921800901002385.

[10] Francisco Sánchez-Bayo, Paul J. van den Brink, and Reiner M. Mann. Eco- logical Impacts of Toxic Chemicals. Bentham Science Publishers, 2012. ISBN 9781608051212. Google-Books-ID: 5oYEMyCTEcwC.

[11] Vikas Singh, Khaiwal Ravindra, Lokesh Sahu, and Ranjeet Sokhi. Trends of atmospheric black carbon concentration over the United Kingdom. Atmospheric En- vironment, 178:148–157, April 2018. ISSN 1352-2310. doi: 10.1016/j.atmosenv. 2018.01.030. URL http://www.sciencedirect.com/science/article/pii/ S135223101830044X.

[12] Prashant Kumar, Allison P.Patton, John L. Durant, and H. Christopher Frey. A review of factors impacting exposure to PM2.5, ultraﬁne particles and black carbon in Asian transport microenvironments. Atmospheric Environment, 187:301–316, 274 August 2018. ISSN 1352-2310. doi: 10.1016/j.atmosenv.2018.05.046. URL http: //www.sciencedirect.com/science/article/pii/S1352231018303509.

[13] Silvia Cerri, Valeria M. Piccolini, Giada Santin, Maria G. Bottone, Sandra A. De Pascali, Danilo Migoni, Paolo Iadarola, Francesco P. Fanizzi, and Graziella Bernocchi. The developmental neurotoxicity study of platinum compounds. Ef- fects of cisplatin versus a novel Pt(II) complex on rat cerebellum. Neurotoxicol- ogy and Teratology, 33(2):273–281, March 2011. ISSN 0892-0362. doi: 10.1016/ j.ntt.2010.09.005. URL http://www.sciencedirect.com/science/article/ pii/S0892036210001698.

[14] Olivia V. Sanderfoot and Tracey Holloway. Air pollution impacts on avian species via inhalation exposure and associated outcomes. Environmental Research Let- ters, 12(8):083002, August 2017. ISSN 1748-9326. doi: 10.1088/1748-9326/ aa8051. URL http://iopscience.iop.org/article/10.1088/1748-9326/ aa8051/meta.

[15] Ilaria Negri, Christian Mavris, Gennaro Di Prisco, Emilio Caprio, and Marco Pel- lecchia. Honey Bees (Apis mellifera, L.) as Active Samplers of Airborne Par- ticulate Matter. PLOS ONE, 10(7):e0132491, July 2015. ISSN 1932-6203. doi: 10.1371/journal.pone.0132491. URL https://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0132491.

[16] Dave Goulson. The demise of the bumblebee in Britain, 2006. URL https:// www.stir.ac.uk/research/hub/publication/4273.

[17] Dave Goulson. Bumblebees: Behaviour, Ecology, and Conservation. OUP Oxford, 2010. ISBN 9780199553068. Google-Books-ID: F6wVDAAAQBAJ.

[18] Ann-Marie Rottler-Hoermann, Stefan Schulz, and Manfred Ayasse. Nest wax triggers worker reproduction in the bumblebee Bombus terrestris. Royal Soci- ety Open Science, 3(1), January 2016. ISSN 2054-5703. doi: 10.1098/rsos.150599. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4736944/.

275 [19] D. Goulson, W. Hughes, L. Derwent, and J. Stout. Colony growth of the bumblebee, Type Bombus terrestris, in improved and conventional agricultural and suburban habitats. Oecologia, 130(2):267–273, January 2002. ISSN 0029-8549, 1432-1939. doi: 10.1007/s004420100803. URL https://link.springer.com/ article/10.1007/s004420100803.

[20] M. J. Duchateau and H. H. W.Velthuis. Development and Reproductive Strategies in Bombus Terrestris Colonies. Behaviour, 107(3):186–207, January 1988. ISSN 1568-539X. doi: 10.1163/156853988X00340. URL http://booksandjournals. brillonline.com/content/journals/10.1163/156853988x00340.

[21] Stephan Wolf and Robin F. A. Moritz. Foraging distance in Bombus terrestris L. (Hymenoptera: Apidae). Api- dologie, 39(4):419–427, July 2008. ISSN 0044-8435, 1297-9678. doi: 10. 1051/apido:2008020. URL https://link.springer.com/article/10.1051/ apido:2008020.

[22] J. B. Free. Insect pollination of crops. Insect pollination of crops., (Ed. 2), 1993. URL https://www.cabdirect.org/cabdirect/abstract/19930233533.

[23] Ian J. Bateman, Georgina M. Mace, Carlo Fezzi, Giles Atkinson, and Kerry Turner. Economic Analysis for Ecosystem Service Assessments. Environmental and Resource Economics, 48(2):177–218, February 2011. ISSN 0924-6460, 1573- 1502. doi: 10.1007/s10640-010-9418-x. URL https://link.springer.com/ article/10.1007/s10640-010-9418-x.

[24] John E. Losey and Mace Vaughan. The Economic Value of Ecologi- cal Services Provided by Insects. BioScience, 56(4):311–323, April 2006. ISSN 0006-3568. doi: 10.1641/0006-3568(2006)56[311:TEVOES]2.0.CO;2. URL http://www.bioone.org/doi/abs/10.1641/0006-3568%282006%2956% 5B311%3ATEVOES%5D2.0.CO%3B2.

276 [25] Paul A De Luca and Mario Vallejo-Marín. What’s the ‘buzz’ about? The ecology and evolutionary signiﬁcance of buzz-pollination. Current Opinion in Plant Biology, 16(4):429–435, August 2013. ISSN 1369-5266. doi: 10.1016/j.pbi. 2013.05.002. URL http://www.sciencedirect.com/science/article/pii/ S1369526613000630.

[26] Shinichi Asada and Masato Ono. Difference in colony development of two Japanese bumblebees, Bombus hypocrita and B. ignitus (Hymenoptera: Api- dae). Applied Entomology and Zoology, 35(4):597–603, 2000. ISSN 0003-6862, 1347-605X. doi: 10.1303/aez.2000.597. URL https://www.jstage.jst.go.jp/ article/aez/35/4/35_4_597/_article/-char/ja/.

[27] Sarah A. Corbet, Ingrid H. Williams, and Juliet L. Osborne. Bees and the Pollina- tion of Crops and Wild Flowers in the European Community. Bee World, 72(2): 47–59, January 1991. ISSN 0005-772X. doi: 10.1080/0005772X.1991.11099079. URL https://doi.org/10.1080/0005772X.1991.11099079.

[28] James D. Thomson and R. C. Plowright. Pollen carryover, nectar rewards, and pollinator behavior with special reference to Diervilla lonicera. Oecologia, 46(1):68–74, July 1980. ISSN 0029-8549, 1432- 1939. doi: 10.1007/BF00346968. URL https://link.springer.com/article/ 10.1007/BF00346968.

[29] Alexandra D. Papanikolaou, Ingolf Kühn, Mark Frenzel, Michael Kuhlmann, Pe- ter Poschlod, Simon G. Potts, Stuart P.M. Roberts, and Oliver Schweiger. Wild bee and ﬂoral diversity co-vary in response to the direct and indirect impacts of land use. Ecosphere, 8(11):e02008, 2017. ISSN 2150-8925. doi: 10.1002/ecs2.2008. URL https://esajournals.onlinelibrary.wiley.com/doi/abs/10.1002/ ecs2.2008.

[30] Dennis vanEngelsdorp, Jay D. Evans, Claude Saegerman, Chris Mullin, Eric Haubruge, Bach Kim Nguyen, Maryann Frazier, Jim Frazier, Diana Cox-Foster, Yanping Chen, Robyn Underwood, David R. Tarpy, and Jeffery S. Pettis. Colony 277 Collapse Disorder: A Descriptive Study. PLOS ONE, 4(8):e6481, August 2009. ISSN 1932-6203. doi: 10.1371/journal.pone.0006481. URL http://journals. plos.org/plosone/article?id=10.1371/journal.pone.0006481.

[31] Dennis vanEngelsdorp, Niko Speybroeck, Jay D. Evans, Bach Kim Nguyen, Chris Mullin, Maryann Frazier, Jim Frazier, Diana Cox-Foster, Yanping Chen, David R. Tarpy, Eric Haubruge, Jeffrey S. Pettis, and Claude Saegerman. Weighing Risk Factors Associated With Bee Colony Collapse Disorder by Classiﬁcation and Re- gression Tree Analysis. Journal of Economic Entomology, 103(5):1517–1523, Oc- tober 2010. ISSN 0022-0493. doi: 10.1603/EC09429. URL https://academic. oup.com/jee/article/103/5/1517/790516.

[32] Nathalie A. Steinhauer, Karen Rennich, Michael E. Wilson, Dewey M. Caron, Eu- gene J. Lengerich, Jeff S. Pettis, Robyn Rose, John A. Skinner, David R. Tarpy, James T. Wilkes, Dennis vanEngelsdorp, and for the Bee Informed Partnership. A national survey of managed honey bee 2012–2013 annual colony losses in the USA: results from the Bee Informed Partnership. Journal of Apicultural Research, 53(1):1–18, January 2014. ISSN 0021-8839. doi: 10.3896/IBRA.1.53.1.01. URL https://doi.org/10.3896/IBRA.1.53.1.01.

[33] Romée van der Zee, Lennard Pisa, Sreten Andonov, Robert Brodschneider, Jean- Daniel Charrière, Róbert Chlebo, Mary F. Coffey, Karl Crailsheim, Bjørn Dahle, Anna Gajda, Alison Gray, Marica M. Drazic, Mariano Higes, Lassi Kauko, Aykut Kence, Meral Kence, Nicola Kezic, Hrisula Kiprijanovska, Jasna Kralj, Preben Kristiansen, Raquel Martin Hernandez, Franco Mutinelli, Bach Kim Nguyen, Christoph Otten, Asli Özkırım, Stephen F.Pernal, Magnus Peterson, Gavin Ram- say, Violeta Santrac, Victoria Soroker, Grazyna˙ Topolska, Aleksandar Uzunov, Flemming Vejsnæs, Shi Wei, and Selwyn Wilkins. Managed honey bee colony losses in Canada, China, Europe, Israel and Turkey, for the winters of 2008–9 and 2009–10. Journal of Apicultural Research, 51(1):100–114, January 2012. ISSN

278 0021-8839. doi: 10.3896/IBRA.1.51.1.12. URL https://doi.org/10.3896/ IBRA.1.51.1.12.

[34] Robert Brodschneider, Alison Gray, Romée van der Zee, Noureddine Adjlane, Valters Brusbardis, Jean-Daniel Charrière, Robert Chlebo, Mary F. Coffey, Karl Crailsheim, Bjørn Dahle, JiˇríDanihlík, Ellen Danneels, Dirk C. de Graaf, Mar- ica Maja Draži´c, Mariia Fedoriak, Ivan Forsythe, Miroljub Golubovski, Ales Gregorc, Urszula Grz˛eda,Ian Hubbuck, Rah¸san ˙Ivgin Tunca, Lassi Kauko, Ole Kilpinen, Justinas Kretavicius, Preben Kristiansen, Maritta Martikkala, Raquel Martín-Hernández, Franco Mutinelli, Magnus Peterson, Christoph Otten, Aslı Ozkirim, Aivar Raudmets, Noa Simon-Delso, Victoria Soroker, Grazyna Topolska, Julien Vallon, Flemming Vejsnæs, and Saskia Woehl. Preliminary analysis of loss rates of honey bee colonies during winter 2015/16 from the COLOSS survey. Jour- nal of Apicultural Research, 55(5):375–378, December 2016. ISSN 0021-8839. doi: 10.1080/00218839.2016.1260240. URL https://doi.org/10.1080/00218839. 2016.1260240.

[35] Parham Georgia. In a race against extinction, rusty patched bumble bee is listed as endangered | U.S. Fish and Wildlife Service Midwest Region, 2017. URL https://www.fws.gov/midwest/news/861.html.

[36] Schmid-Hempel Regula, Eckhardt Michael, Goulson David, Heinzmann Daniel, Lange Carlos, Plischuk Santiago, Escudero Luisa R., Salathé Rahel, Scriven Jes- sica J., Schmid-Hempel Paul, and Boots Mike. The invasion of southern South America by imported bumblebees and associated parasites. Journal of Animal Ecology, 83(4):823–837, January 2014. ISSN 0021-8790. doi: 10.1111/1365-2656. 12185. URL https://besjournals.onlinelibrary.wiley.com/doi/full/ 10.1111/1365-2656.12185.

[37] Paul H. Williams and Juliet L. Osborne. Bumblebee vulnerability and conservation world-wide. Apidologie, 40(3):367–387, May 2009. ISSN 0044-8435, 1297- 9678. doi: 10.1051/apido/2009025. URL http://dx.doi.org/10.1051/apido/

279 2009025.

[38] Jennifer C. Grixti, Lisa T. Wong, Sydney A. Cameron, and Colin Favret. Decline of bumble bees (Bombus) in the North American Midwest. Biological Conser- vation, 142(1):75–84, January 2009. ISSN 0006-3207. doi: 10.1016/j.biocon. 2008.09.027. URL http://www.sciencedirect.com/science/article/pii/ S0006320708003698.

[39] Zhenghua Xie, Paul H. Williams, and Ya Tang. The effect of grazing on bumblebees in the high rangelands of the eastern Tibetan Plateau of Sichuan. Journal of Insect Conservation, 12(6):695–703, December 2008. ISSN 1366-638X, 1572- 9753. doi: 10.1007/s10841-008-9180-3. URL https://link.springer.com/ article/10.1007/s10841-008-9180-3.

[40] Andrzej Kosior, Waldemar Celary, Paweł Olejniczak, Jan Fijał, Wieslaw Król, Wojciech Solarz, and Piotr Płonka. The decline of the bumble bees and cuckoo bees (Hymenoptera: Apidae: Bombini) of Western and Central Europe. Oryx, 41(1):79–88, January 2007. ISSN 1365-3008, 0030-6053. doi: 10.1017/ S0030605307001597. URL https://scinapse.io/papers/2133419583.

[41] Michelle L. Hladik, Anson R. Main, and Dave Goulson. Environmental Risks and Challenges Associated with Neonicotinoid Insecticides. Environmental Science & Technology, 52(6):3329–3335, March 2018. ISSN 0013-936X. doi: 10.1021/acs. est.7b06388. URL https://doi.org/10.1021/acs.est.7b06388.

[42] Simon G. Potts, Jacobus C. Biesmeijer, Claire Kremen, Peter Neumann, Oliver Schweiger, and William E. Kunin. Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution, 25(6):345–353, June 2010. ISSN 0169- 5347. doi: 10.1016/j.tree.2010.01.007. URL http://www.sciencedirect.com/ science/article/pii/S0169534710000364.

280 [43] Mark J. F. Brown and Robert J. Paxton. The conservation of bees: a global perspective. Apidologie, 40(3):410–416, May 2009. ISSN 0044-8435, 1297-9678. doi: 10.1051/apido/2009019. URL http://dx.doi.org/10.1051/apido/2009019.

[44] Matt McGrath. Pesticide link to long-term bee decline. BBC News, August 2016. URL http://www.bbc.co.uk/news/science-environment-37089385.

[45] Ivan Meeus, Matti Pisman, Guy Smagghe, and Niels Piot. Interaction effects of different drivers of wild bee decline and their inﬂuence on host–pathogen dynamics. Current Opinion in Insect Science, 26:136–141, April 2018. ISSN 2214- 5745. doi: 10.1016/j.cois.2018.02.007. URL http://www.sciencedirect.com/ science/article/pii/S2214574517301803.

[46] C. Ruth Archer, Christian W. W. Pirk, Geraldine A. Wright, and Sue W. Nicol- son. Nutrition affects survival in African honeybees exposed to interacting stressors. Functional Ecology, 28(4):913–923, 2013. ISSN 1365-2435. doi: 10.1111/ 1365-2435.12226. URL https://besjournals.onlinelibrary.wiley.com/ doi/abs/10.1111/1365-2435.12226.

[47] Rebecca K. Tonietto and Daniel J. Larkin. Habitat restoration beneﬁts wild bees: A meta-analysis. Journal of Applied Ecology, 55(2):582–590, 2017. ISSN 1365-2664. doi: 10.1111/1365-2664.13012. URL https://besjournals. onlinelibrary.wiley.com/doi/abs/10.1111/1365-2664.13012.

[48] Inka Lusebrink, Robbie D. Girling, Emily Farthing, Tracey A. Newman, Chris W. Jackson, and Guy M. Poppy. The Effects of Diesel Exhaust Pollution on Floral Volatiles and the Consequences for Honey Bee Olfaction. Journal of Chemi- cal Ecology, 41(10):904–912, October 2015. ISSN 0098-0331, 1573-1561. doi: 10.1007/s10886-015-0624-4. URL https://link.springer.com/article/10. 1007/s10886-015-0624-4.

[49] Thomas James Wood and Dave Goulson. The environmental risks of neonicotinoid pesticides: a review of the evidence post 2013. Environmental Science

281 and Pollution Research, 24(21):17285–17325, July 2017. ISSN 0944-1344, 1614- 7499. doi: 10.1007/s11356-017-9240-x. URL https://link.springer.com/ article/10.1007/s11356-017-9240-x.

[50] Mickaël Henry, Matthias A. Becher, Juliet L. Osborne, Peter J. Kennedy, Pier- rick Aupinel, Vincent Bretagnolle, François Brun, Volker Grimm, Juliane Horn, and Fabrice Requier. Predictive systems models can help elucidate bee declines driven by multiple combined stressors. Apidologie, 48(3):328–339, May 2017. ISSN 0044-8435, 1297-9678. doi: 10.1007/s13592-016-0476-0. URL https://link.springer.com/article/10.1007/s13592-016-0476-0.

[51] Aline Fauser, Christoph Sandrock, Peter Neumann, and Ben M. Sadd. Neonicoti- noids override a parasite exposure impact on hibernation success of a key bumblebee pollinator. Ecological Entomology, 42(3):306–314, June 2017. ISSN 1365- 2311. doi: 10.1111/een.12385. URL https://onlinelibrary.wiley.com/doi/ abs/10.1111/een.12385.

[52] Dave Goulson, Elizabeth Nicholls, Cristina Botías, and Ellen L. Rotheray. Bee declines driven by combined stress from parasites, pesticides, and lack of ﬂow- ers. Science, 347(6229):1255957, March 2015. ISSN 0036-8075, 1095-9203. doi: 10.1126/science.1255957. URL http://science.sciencemag.org/content/ 347/6229/1255957.

[53] Dave Goulson. REVIEW: An overview of the environmental risks posed by neonicotinoid insecticides. Journal of Applied Ecology, 50(4):977–987, 2013. ISSN 1365-2664. doi: 10.1111/1365-2664.12111. URL https://besjournals. onlinelibrary.wiley.com/doi/abs/10.1111/1365-2664.12111.

[54] John E. Casida. Neonicotinoids and Other Insect Nicotinic Receptor Competitive Modulators: Progress and Prospects. Annual Review of Entomology, 63(1):125– 144, 2018. doi: 10.1146/annurev-ento-020117-043042. URL https://doi.org/ 10.1146/annurev-ento-020117-043042.

282 [55] Kasie Raymann, Erick V. S. Motta, Catherine Girard, Ian M. Riddington, Jor- dan A. Dinser, and Nancy A. Moran. Imidacloprid decreases honey bee survival but does not affect the gut microbiome. Applied and Environmental Microbi- ology, pages AEM.00545–18, April 2018. ISSN 0099-2240, 1098-5336. doi: 10. 1128/AEM.00545-18. URL http://aem.asm.org/content/early/2018/04/ 16/AEM.00545-18.

[56] Susan C. Anenberg, Joel Schwartz, Drew Shindell, Markus Amann, Greg Faluvegi, Zbigniew Klimont, Greet Janssens-Maenhout, Luca Pozzoli, Rita Van Dingenen, Elisabetta Vignati, Lisa Emberson, Nicholas Z. Muller, J. Jason West, Martin Williams, Volodymyr Demkine, W. Kevin Hicks, Johan Kuylenstierna, Frank Raes, and Veerabhadran Ramanathan. Global air quality and health co-beneﬁts of mit- igating near-term climate change through methane and black carbon emission controls. Environmental Health Perspectives, 120(6):831–839, June 2012. ISSN 1552-9924. doi: 10.1289/ehp.1104301.

[57] C. A. Masiello. New directions in black carbon organic geochemistry. Ma- rine Chemistry, 92(1):201–213, December 2004. ISSN 0304-4203. doi: 10.1016/ j.marchem.2004.06.043. URL http://www.sciencedirect.com/science/ article/pii/S0304420304002051.

[58] Sunghwan Kim, Louis A. Kaplan, Ronald Benner, and Patrick G. Hatcher. Hydrogen-deﬁcient molecules in natural riverine water samples—evidence for the existence of black carbon in DOM. Marine Chemistry, 92(1):225–234, De- cember 2004. ISSN 0304-4203. doi: 10.1016/j.marchem.2004.06.042. URL http: //www.sciencedirect.com/science/article/pii/S0304420304002075.

[59] Michael Schmidt, Jan O. Skjemstad, Claudia Czimczik, Bruno Glaser, Ken M. Prentice, Yves Gelinas, and T.A.J. Kuhlbusch. Comparative analysis of black carbon in soils. Global Biogeochemical Cycles, 15, 2001. doi: 10.1029/ 2000GB001284.

283 [60] Aron Stubbins, Robert G. M. Spencer, Hongmei Chen, Patrick G. Hatcher, Ken- neth Mopper, Peter J. Hernes, Vincent L. Mwamba, Arthur M. Mangangu, Jose N. Wabakanghanzi, and Johan Six. Illuminated darkness: Molecular signatures of Congo River dissolved organic matter and its photochemical alteration as revealed by ultrahigh precision mass spectrometry. Limnology and Oceanogra- phy, 55(4):1467–1477, 2010. ISSN 1939-5590. doi: 10.4319/lo.2010.55.4.1467. URL https://aslopubs.onlinelibrary.wiley.com/doi/abs/10.4319/lo. 2010.55.4.1467.

[61] T. C. Bond, S. J. Doherty, D. W. Fahey, P. M. Forster, T. Berntsen, B. J. DeAn- gelo, M. G. Flanner, S. Ghan, B. Kärcher, D. Koch, S. Kinne, Y. Kondo, P. K. Quinn, M. C. Saroﬁm, M. G. Schultz, M. Schulz, C. Venkataraman, H. Zhang, S. Zhang, N. Bellouin, S. K. Guttikunda, P.K. Hopke, M. Z. Jacobson, J. W. Kaiser, Z. Klimont, U. Lohmann, J. P. Schwarz, D. Shindell, T. Storelvmo, S. G. War- ren, and C. S. Zender. Bounding the role of black carbon in the climate system: A scientiﬁc assessment. Journal of Geophysical Research: Atmospheres, 118 (11):5380–5552, 2013. ISSN 2169-8996. doi: 10.1002/jgrd.50171. URL https: //agupubs.onlinelibrary.wiley.com/doi/abs/10.1002/jgrd.50171.

[62] Michael. Brauer. Health Effects of Transport-related Air Pollution - ProQuest, 2006. URL https://search.proquest.com/openview/ f6ad94c3c78a42e3b19f5f852baf673c/1?pq-origsite=gscholar&cbl= 47649.

[63] Francine Laden, Joel Schwartz, Frank E. Speizer, and Douglas W.Dockery. Reduc- tion in Fine Particulate Air Pollution and Mortality. American Journal of Respi- ratory and Critical Care Medicine, 173(6):667–672, March 2006. ISSN 1073-449X. doi: 10.1164/rccm.200503-443OC. URL https://www.atsjournals.org/doi/ abs/10.1164/rccm.200503-443OC.

[64] B. Liang, J. Lehmann, D. Solomon, J. Kinyangi, J. Grossman, B. O’Neill, J. O. Skjemstad, J. Thies, F. J. Luizão, J. Petersen, and E. G. Neves. Black Carbon

284 Increases Cation Exchange Capacity in Soils. Soil Science Society of America Journal, 70(5):1719–1730, September 2006. ISSN 1435-0661. doi: 10.2136/ sssaj2005.0383. URL https://dl.sciencesocieties.org/publications/ sssaj/abstracts/70/5/1719.

[65] D. A Grantz, J. H. B Garner, and D. W Johnson. Ecological effects of particulate matter. Environment International, 29(2):213–239, June 2003. ISSN 0160-4120. doi: 10.1016/S0160-4120(02)00181-2. URL http://www.sciencedirect.com/ science/article/pii/S0160412002001812.

[66] Shane J. K. Hussey, Joanne Purves, Natalie Allcock, Vitor E. Fernandes, Paul S. Monks, Julian M. Ketley, Peter W. Andrew, and Julie A. Morrissey. Air pollution alters Staphylococcus aureus and Streptococcus pneumoniae bioﬁlms, antibiotic tolerance and colonisation. Environmental Microbiology, 19(5):1868– 1880, May 2017. ISSN 1462-2920. doi: 10.1111/1462-2920.13686. URL https: //onlinelibrary.wiley.com/doi/abs/10.1111/1462-2920.13686.

[67] Kazuhiko Matsuda, Steven D. Buckingham, Daniel Kleier, James J. Rauh, Marta Grauso, and David B. Sattelle. Neonicotinoids: insecticides acting on insect nicotinic acetylcholine receptors. Trends in Pharmacological Sci- ences, 22(11):573–580, November 2001. ISSN 0165-6147. doi: 10.1016/ S0165-6147(00)01820-4. URL http://www.sciencedirect.com/science/ article/pii/S0165614700018204.

[68] Motohiro Tomizawa and John E. Casida. Selective toxicity of neonicotinoids attributable to speciﬁcity of insect and mammalian nicotinic receptors. Annual Review of Entomology, 48:339–364, 2003. ISSN 0066-4170. doi: 10.1146/annurev. ento.48.091801.112731.

[69] Emiliane Taillebois, Alison Cartereau, Andrew K. Jones, and Steeve H. Thany. Neonicotinoid insecticides mode of action on insect nicotinic acetylcholine receptors using binding studies. Pesticide Biochemistry and Physiology, 151:59–66,

285 October 2018. ISSN 0048-3575. doi: 10.1016/j.pestbp.2018.04.007. URL http: //www.sciencedirect.com/science/article/pii/S0048357517306296.

[70] Christopher Moffat, Stephen T. Buckland, Andrew J. Samson, Robin McArthur, Victor Chamosa Pino, Karen A. Bollan, Jeffrey T.-J. Huang, and Christopher N. Connolly. Neonicotinoids target distinct nicotinic acetylcholine receptors and neurons, leading to differential risks to bumblebees. Scientiﬁc Reports, 6:24764, April 2016. ISSN 2045-2322. doi: 10.1038/srep24764. URL https://www. nature.com/articles/srep24764.

[71] Á. Vehovszky, A. Farkas, A. Ács, O. Stoliar, A. Székács, M. Mörtl, and J. Gy˝ori. Neonicotinoid insecticides inhibit cholinergic neurotransmission in a mollus- can (Lymnaea stagnalis) nervous system. Aquatic Toxicology, 167:172–179, Oc- tober 2015. ISSN 0166-445X. doi: 10.1016/j.aquatox.2015.08.009. URL http: //www.sciencedirect.com/science/article/pii/S0166445X15300357.

[72] Motohiro Tomizawa. Chapter Two - Chemical Biology of the Nicotinic In- secticide Receptor. In Ephraim Cohen, editor, Advances in Insect Physiol- ogy, volume 44 of Target Receptors in the Control of Insect Pests: Part I, pages 63–99. Academic Press, January 2013. doi: 10.1016/B978-0-12-394389-7. 00002-8. URL http://www.sciencedirect.com/science/article/pii/ B9780123943897000028.

[73] I. Yamamoto and J. E. Casida, editors. Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor. Springer Japan, 1999. ISBN 9784431702139. URL //www. springer.com/gb/book/9784431702139.

[74] Peter Jeschke, Ralf Nauen, Michael Schindler, and Alfred Elbert. Overview of the status and global strategy for neonicotinoids. Journal of Agricultural and Food Chemistry, 59(7):2897–2908, April 2011. ISSN 1520-5118. doi: 10.1021/jf101303g.

286 [75] Ignaz J. Buerge, Astrid Bächli, Werner E. Heller, Martina Keller, and Thomas Poiger. Environmental Behavior of the Chiral Herbicide Haloxyfop. 2. Un- changed Enantiomer Composition in Blackgrass (Alopecurus myosuroides) and Garden Cress (Lepidium sativum). Journal of Agricultural and Food Chemistry, 63(10):2591–2596, March 2015. ISSN 0021-8561. doi: 10.1021/jf505242f. URL https://doi.org/10.1021/jf505242f.

[76] Leonardo C. Magalhaes, Thomas E. Hunt, and Blair D. Siegfried. Efﬁcacy of Neonicotinoid Seed Treatments to Reduce Soybean Aphid Populations Under Field and Controlled Conditions in Nebraska. Journal of Economic Entomol- ogy, 102(1):187–195, 2009. ISSN 0022-0493. doi: 10.1603/029.102.0127. URL http://www.bioone.org/doi/abs/10.1603/029.102.0127.

[77] Bernadette M. Mach, Svetlana Bondarenko, and Daniel A. Potter. Uptake and dis- sipation of neonicotinoid residues in nectar and foliage of systemically treated woody landscape plants. Environmental Toxicology and Chemistry, 37(3):860– 870, March 2018. ISSN 1552-8618. doi: 10.1002/etc.4021.

[78] Cristina Botías, Arthur David, Julia Horwood, Alaa Abdul-Sada, Elizabeth Nicholls, Elizabeth Hill, and Dave Goulson. Neonicotinoid Residues in Wild- ﬂowers, a Potential Route of Chronic Exposure for Bees. Environmental Sci- ence & Technology, 49(21):12731–12740, November 2015. ISSN 0013-936X. doi: 10.1021/acs.est.5b03459. URL https://doi.org/10.1021/acs.est.5b03459.

[79] Peter Pollak. Fine Chemicals: The Industry and the Business. John Wiley & Sons, March 2011. ISBN 9781118002223. Google-Books-ID: yuxJGasHvkoC.

[80] European Food Safety Authority. Neonicotinoids: risks to bees conﬁrmed, Febru- ary 2018. URL https://www.efsa.europa.eu/en/press/news/180228.

[81] National Farmers Union. NFU reaction to neonicotinoids announcement, September 2017. URL https://www.nfuonline.com:443/cross-sector/

287 environment/bees-and-pollinators/bees-and-pollinators-news/ nfu-reaction-to-neonicotinoids-announcement/.

[82] Pubchem. Imidacloprid, 2018. URL https://pubchem.ncbi.nlm.nih.gov/ compound/86418.

[83] Laura Jesse. Homeowner Imidacloprid Recommendations for Edible Crops | Horticulture and Home Pest News, 2012. URL https://hortnews.extension. iastate.edu/2012/3-21/imidacloprid.html.

[84] fera. PESTICIDE USAGE STATISTICS - PUS STATS. URL https://secure.fera. defra.gov.uk/pusstats/mygraphindex.cfm.

[85] Jiangong Jiang, Dicheng Ma, Nan Zou, Xin Yu, Zhengqun Zhang, Feng Liu, and Wei Mu. Concentrations of imidacloprid and thiamethoxam in pollen, nectar and leaves from seed-dressed cotton crops and their potential risk to honeybees (Apis mellifera L.). Chemosphere, 201:159–167, June 2018. ISSN 0045-6535. doi: 10.1016/j.chemosphere.2018.02.168. URL http://www.sciencedirect.com/ science/article/pii/S0045653518303801.

[86] Christy A. Morrissey, Pierre Mineau, James H. Devries, Francisco Sanchez-Bayo, Matthias Liess, Michael C. Cavallaro, and Karsten Liber. Neonicotinoid contamination of global surface waters and associated risk to aquatic invertebrates: A review. Environment International, 74:291–303, January 2015. ISSN 0160- 4120. doi: 10.1016/j.envint.2014.10.024. URL http://www.sciencedirect. com/science/article/pii/S0160412014003183.

[87] Tessa C. Van Dijk, Marja A. Van Staalduinen, and Jeroen P.Van der Sluijs. Macro- Invertebrate Decline in Surface Water Polluted with Imidacloprid. PLOS ONE, 8(5):e62374, May 2013. ISSN 1932-6203. doi: 10.1371/journal.pone.0062374. URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0062374.

288 [88] buglife. Heavy Neonicotinoid Insecticide Contamina- tion Damaging British Rivers | Buglife, 2017. URL https://www.buglife.org.uk/news-and-events/news/ heavy-neonicotinoid-insecticide-contamination-damaging-british-rivers.

[89] Francisco Sanchez-Bayo and Koichi Goka. Pesticide Residues and Bees – A Risk Assessment. PLOS ONE, 9(4):e94482, April 2014. ISSN 1932-6203. doi: 10.1371/ journal.pone.0094482. URL http://journals.plos.org/plosone/article? id=10.1371/journal.pone.0094482.

[90] Richard Schmuck, Ralf Schöning, Andreas Stork, and Oliver Schramel. Risk posed to honeybees (Apis mellifera L, Hymenoptera) by an imidacloprid seed dressing of sunﬂowers. Pest Management Science, 57(3):225–238, 2001. ISSN 1526-4998. doi: 10.1002/ps.270. URL https://onlinelibrary.wiley.com/ doi/abs/10.1002/ps.270.

[91] J. M. Bonmatin, I. Moineau, R. Charvet, C. Fleche, M. E. Colin, and E. R. Bengsch. A LC/APCI-MS/MS method for analysis of imidacloprid in soils, in plants, and in pollens. Analytical Chemistry, 75(9):2027–2033, May 2003. ISSN 0003-2700. doi: 10.1021/ac020600b.

[92] François M. Laurent and Estelle Rathahao. Distribution of [(14)C]imidacloprid in sunﬂowers (Helianthus annuus L.) following seed treatment. Journal of Agri- cultural and Food Chemistry, 51(27):8005–8010, December 2003. ISSN 0021- 8561. doi: 10.1021/jf034310n. URL https://pubs.acs.org/doi/abs/10. 1021/jf034310n.

[93] Ben A. Woodcock, Lucy Ridding, Stephen N. Freeman, M. Gloria Pereira, Dar- ren Sleep, John Redhead, David Aston, Norman L. Carreck, Richard F. Shore, James M. Bullock, Matthew S. Heard, and Richard F. Pywell. Neonicotinoid residues in UK honey despite European Union moratorium. PLOS ONE, 13(1): e0189681, January 2018. ISSN 1932-6203. doi: 10.1371/journal.pone.0189681.

289 URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0189681.

[94] Christopher N. Connolly. Nerve agents in honey. Science, 358(6359):38–39, Oc- tober 2017. ISSN 0036-8075, 1095-9203. doi: 10.1126/science.aao6000. URL http://science.sciencemag.org/content/358/6359/38.

[95] John D. Stark, Paul C. Jepson, and Daniel F. Mayer. Limitations to Use of Top- ical Toxicity Data for Predictions of Pesticide Side Effects in the Field. Jour- nal of Economic Entomology, 88(5):1081–1088, October 1995. ISSN 0022-0493. doi: 10.1093/jee/88.5.1081. URL https://academic.oup.com/jee/article/ 88/5/1081/2216322.

[96] Séverine Suchail, David Guez, and Luc P. Belzunces. Characteristics of imidacloprid toxicity in two Apis mellifera subspecies. Environmental Toxicol- ogy and Chemistry, 19(7):1901–1905, 2000. ISSN 1552-8618. doi: 10.1002/etc. 5620190726. URL https://setac.onlinelibrary.wiley.com/doi/abs/10. 1002/etc.5620190726.

[97] Takao Iwasa, Naoki Motoyama, John T. Ambrose, and R. Michael Roe. Mech- anism for the differential toxicity of neonicotinoid insecticides in the honey bee, Apis mellifera. Crop Protection, 23(5):371–378, May 2004. ISSN 0261- 2194. doi: 10.1016/j.cropro.2003.08.018. URL http://www.sciencedirect. com/science/article/pii/S0261219403002308.

[98] S. Suchail, D. Guez, and L. P.Belzunces. Discrepancy between acute and chronic toxicity induced by imidacloprid and its metabolites in Apis mellifera. Environ- mental Toxicology and Chemistry, 20(11):2482–2486, November 2001. ISSN 0730- 7268.

[99] Francisco Sánchez-Bayo, Luc Belzunces, and Jean-Marc Bonmatin. Lethal and sublethal effects, and incomplete clearance of ingested imidacloprid in honey bees (Apis mellifera). Ecotoxicology, 26

290 (9):1199–1206, November 2017. ISSN 0963-9292, 1573-3017. doi: 10.1007/ s10646-017-1845-9. URL https://link.springer.com/article/10.1007/ s10646-017-1845-9.

[100] Tjeerd Blacquière, Guy Smagghe, Cornelis A. M. van Gestel, and Veerle Mom- maerts. Neonicotinoids in bees: a review on concentrations, side-effects and risk assessment. Ecotoxicology (London, England), 21(4):973–992, May 2012. ISSN 0963-9292. doi: 10.1007/s10646-012-0863-x. URL https://www.ncbi. nlm.nih.gov/pmc/articles/PMC3338325/.

[101] Helen M. Thompson, Selwyn Wilkins, Sarah Harkin, Sarah Milner, and Keith FA Walters. Neonicotinoids and bumblebees (Bombus terrestris): effects on nectar consumption in individual workers. Pest Management Science, 71(7):946–950, 2014. ISSN 1526-4998. doi: 10.1002/ps.3868. URL https://onlinelibrary. wiley.com/doi/abs/10.1002/ps.3868.

[102] Elizabeth Nicholls, Robert Fowler, Jeremy E. Niven, James D. Gilbert, and Dave Goulson. Larval exposure to ﬁeld-realistic concentrations of clothianidin has no effect on development rate, over-winter survival or adult metabolic rate in a solitary bee, Osmia bicornis. PeerJ, 5, June 2017. ISSN 2167-8359. doi: 10.7717/peerj. 3417. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5480390/.

[103] Abdessalam Kacimi El Hassani, Matthieu Dacher, Vincent Gary, Michel Lam- bin, Monique Gauthier, and Catherine Armengaud. Effects of Sublethal Doses of Acetamiprid and Thiamethoxam on the Behavior of the Honeybee (Apis mellifera). Archives of Environmental Con- tamination and Toxicology, 54(4):653–661, May 2008. ISSN 0090-4341, 1432- 0703. doi: 10.1007/s00244-007-9071-8. URL https://link.springer.com/ article/10.1007/s00244-007-9071-8.

[104] Michael C. Newman. Fundamentals of Ecotoxicology, Third Edition. CRC Press, 2009. ISBN 9781439883129. Google-Books-ID: sg_SBQAAQBAJ.

291 [105] Judy Wu-Smart and Marla Spivak. Sub-lethal effects of dietary neonicotinoid insecticide exposure on honey bee queen fecundity and colony development. Sci- entiﬁc Reports, 6:32108, August 2016. ISSN 2045-2322. doi: 10.1038/srep32108. URL https://www.nature.com/articles/srep32108.

[106] Jamison Scholer and Vera Krischik. Chronic exposure of imidacloprid and clothianidin reduce queen survival, foraging, and nectar storing in colonies of Bombus impatiens. PloS One, 9(3):e91573, 2014. ISSN 1932-6203. doi: 10.1371/journal. pone.0091573.

[107] J. N. Tasei, G. Ripault, and E. Rivault. Hazards of imidacloprid seed coating to Bombus terrestris (Hymenoptera: Apidae) when applied to sunﬂower. Journal of Economic Entomology, 94(3):623–627, June 2001. ISSN 0022-0493.

[108] Richard J. Gill, Oscar Ramos-Rodriguez, and Nigel E. Raine. Combined pesticide exposure severely affects individual- and colony-level traits in bees. Nature, 491 (7422):105–108, November 2012. ISSN 1476-4687. doi: 10.1038/nature11585. URL https://www.nature.com/articles/nature11585.

[109] Maria Arena and Fabio Sgolastra. A meta-analysis comparing the sensitivity of bees to pesticides. Ecotoxicology (London, England), 23(3):324–334, April 2014. ISSN 1573-3017. doi: 10.1007/s10646-014-1190-1.

[110] Luc P. Belzunces, Sylvie Tchamitchian, and Jean-Luc Brunet. Neural effects of insecticides in the honey bee. Apidologie, 43(3):348–370, May 2012. ISSN 0044-8435, 1297-9678. doi: 10.1007/s13592-012-0134-0. URL https://link. springer.com/article/10.1007/s13592-012-0134-0.

[111] Yassine Aliouane, Adessalam K. El Hassani, Vincent Gary, Catherine Armengaud, Michel Lambin, and Monique Gauthier. Subchronic exposure of honeybees to sublethal doses of pesticides: effects on behavior. Environmental Toxicology and Chemistry, 28(1):113–122, January 2009. ISSN 0730-7268. doi: 10.1897/08-110.1.

292 [112] Simone Tosi, Giovanni Burgio, and James C. Nieh. A common neonicotinoid pesticide, thiamethoxam, impairs honey bee ﬂight ability. Scientiﬁc Reports, 7 (1):1201, April 2017. ISSN 2045-2322. doi: 10.1038/s41598-017-01361-8. URL https://www.nature.com/articles/s41598-017-01361-8.

[113] Axel Decourtye and James Devillers. Ecotoxicity of Neonicotinoid Insecti- cides to Bees. In Insect Nicotinic Acetylcholine Receptors, Advances in Ex- perimental Medicine and Biology, pages 85–95. Springer, New York, NY, 2010. ISBN 9781441964441 9781441964458. doi: 10.1007/978-1-4419-6445-8_8. URL https://link.springer.com/chapter/10.1007/978-1-4419-6445-8_8.

[114] Axel Decourtye, James Devillers, Sophie Cluzeau, Mercedes Charreton, and Minh-Hà Pham-Delègue. Effects of imidacloprid and deltamethrin on associa- tive learning in honeybees under semi-ﬁeld and laboratory conditions. Eco- toxicology and Environmental Safety, 57(3):410–419, March 2004. ISSN 0147- 6513. doi: 10.1016/j.ecoenv.2003.08.001. URL http://www.sciencedirect. com/science/article/pii/S0147651303001477.

[115] Veerle Mommaerts, Soﬁe Reynders, Jana Boulet, Linde Besard, Guido Sterk, and Guy Smagghe. Risk assessment for side-effects of neonicotinoids against bumblebees with and without impairing foraging behavior. Ecotoxicology, 19(1):207, January 2010. ISSN 0963-9292, 1573-3017. doi: 10.1007/s10646-009-0406-2. URL https://link.springer.com/article/10.1007/s10646-009-0406-2.

[116] Gemma L. Baron, Nigel E. Raine, and Mark J. F. Brown. General and species- speciﬁc impacts of a neonicotinoid insecticide on the ovary development and feeding of wild bumblebee queens. Proc. R. Soc. B, 284(1854):20170123, May 2017. ISSN 0962-8452, 1471-2954. doi: 10.1098/rspb.2017.0123. URL http: //rspb.royalsocietypublishing.org/content/284/1854/20170123.

[117] R. Menzel and U. Müller. Learning and Memory in Honeybees: From Behav- ior to Neural Substrates. Annual Review of Neuroscience, 19(1):379–404, March

293 1996. ISSN 0147-006X. doi: 10.1146/annurev.ne.19.030196.002115. URL https: //www.annualreviews.org/doi/10.1146/annurev.ne.19.030196.002115.

[118] Dara A. Stanley, Karen E. Smith, and Nigel E. Raine. Bumblebee learning and memory is impaired by chronic exposure to a neonicotinoid pesticide. Scientiﬁc Reports, 5:16508, November 2015. ISSN 2045-2322. doi: 10.1038/srep16508. URL https://www.nature.com/articles/srep16508.

[119] Sally M. Williamson, Daniel D. Baker, and Geraldine A. Wright. Acute exposure to a sublethal dose of imidacloprid and coumaphos enhances olfactory learning and memory in the honeybee Apis mellifera. Invertebrate Neuroscience, 13(1):63–70, June 2013. ISSN 1354-2516, 1439-1104. doi: 10.1007/s10158-012-0144-7. URL https://link. springer.com/article/10.1007/s10158-012-0144-7.

[120] M. Lambin, C. Armengaud, S. Raymond, and M. Gauthier. Imidacloprid- induced facilitation of the proboscis extension reﬂex habituation in the honeybee. Archives of Insect Biochemistry and Physiology, 48(3):129–134, 2001. ISSN 1520-6327. doi: 10.1002/arch.1065. URL https://onlinelibrary.wiley.com/ doi/abs/10.1002/arch.1065.

[121] Hannah Feltham, Kirsty Park, and Dave Goulson. Field realistic doses of pesticide imidacloprid reduce bumblebee pollen foraging efﬁciency. Ecotoxicol- ogy, 23(3):317–323, April 2014. ISSN 0963-9292, 1573-3017. doi: 10.1007/ s10646-014-1189-7. URL https://link.springer.com/article/10.1007/ s10646-014-1189-7.

[122] Richard J. Gill and Nigel E. Raine. Chronic impairment of bumblebee natural foraging behaviour induced by sublethal pesticide exposure. Functional Ecol- ogy, 28(6):1459–1471, 2014. ISSN 1365-2435. doi: 10.1111/1365-2435.12292. URL https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/ 1365-2435.12292.

294 [123] Annely Brandt, Anna Gorenﬂo, Reinhold Siede, Marina Meixner, and Ralph Büchler. The neonicotinoids thiacloprid, imidacloprid, and clothianidin affect the immunocompetence of honey bees (Apis mellifera L.). Journal of Insect Physiology, 86:40–47, March 2016. ISSN 0022-1910. doi: 10.1016/j.jinsphys. 2016.01.001. URL http://www.sciencedirect.com/science/article/pii/ S0022191016300014.

[124] Simone Tosi, James C. Nieh, Fabio Sgolastra, Riccardo Cabbri, and Piotr Medrzy- cki. Neonicotinoid pesticides and nutritional stress synergistically reduce survival in honey bees. Proceedings. Biological Sciences, 284(1869), December 2017. ISSN 1471-2954. doi: 10.1098/rspb.2017.1711.

[125] Geoffrey R. Williams, Aline Troxler, Gina Retschnig, Kaspar Roth, Orlando Yañez, Dave Shutler, Peter Neumann, and Laurent Gauthier. Neonicotinoid pesticides severely affect honey bee queens. Scientiﬁc Reports, 5:14621, October 2015. ISSN 2045-2322. doi: 10.1038/srep14621. URL https://www.nature.com/ articles/srep14621.

[126] Ivana Tlak Gajger, Martina Sakaˇc, and Aleš Gregorc. Impact of Thi- amethoxam on Honey Bee Queen (Apis mellifera car- nica) Reproductive Morphology and Physiology. Bulletin of Envi- ronmental Contamination and Toxicology, 99(3):297–302, September 2017. ISSN 0007-4861, 1432-0800. doi: 10.1007/s00128-017-2144-0. URL https://link. springer.com/article/10.1007/s00128-017-2144-0.

[127] Lina De Smet, Fani Hatjina, Pavlos Ioannidis, Anna Hamamtzoglou, Karel Schoonvaere, Frédéric Francis, Ivan Meeus, Guy Smagghe, and Dirk C. de Graaf. Stress indicator gene expression proﬁles, colony dynamics and tissue development of honey bees exposed to sub-lethal doses of imidacloprid in laboratory and ﬁeld experiments. PLOS ONE, 12(2):e0171529, February 2017. ISSN 1932- 6203. doi: 10.1371/journal.pone.0171529. URL http://journals.plos.org/ plosone/article?id=10.1371/journal.pone.0171529.

295 [128] Andrzej Ciereszko, Jerzy Wilde, Grzegorz J. Dietrich, Maciej Siuda, Beata B ˛ak,Syl- wia Judycka, and Halina Karol. Sperm parameters of honeybee drones exposed to imidacloprid. Apidologie, 48(2):211–222, March 2017. ISSN 0044-8435, 1297- 9678. doi: 10.1007/s13592-016-0466-2. URL https://link.springer.com/ article/10.1007/s13592-016-0466-2.

[129] Philippe J. Berny, Florence Buronfosse, Bernadette Videmann, and Thierry Buronfosse. Evaluation of the Toxicity of Imidacloprid in Wild Birds. a New High Performance Thin Layer Chromatography (hptlc) Method for the Analy- sis of Liver and Crop Samples in Suspected Poisoning Cases. Journal of Liquid Chromatography & Related Technologies, 22(10):1547–1559, January 1999. ISSN 1082-6076. doi: 10.1081/JLC-100101750. URL https://doi.org/10.1081/ JLC-100101750.

[130] Ehsan H. Abu Zeid, Rasha T. M. Alam, Sozan A. Ali, and Mohamed Y. Hen- dawi. Dose-related impacts of imidacloprid oral intoxication on brain and liver of rock pigeon (Columba livia domestica), residues analysis in different organs. Ecotoxicology and Environmental Safety, 167:60–68, January 2019. ISSN 0147- 6513. doi: 10.1016/j.ecoenv.2018.09.121. URL http://www.sciencedirect. com/science/article/pii/S0147651318309965.

[131] Aline Fernanda Catae, Thaisa Cristina Roat, Regiane Alves De Oliveira, Roberta CornéLio Ferreira Nocelli, and Osmar Malaspina. Cytotoxic effects of thiamethoxam in the midgut and malpighian tubules of Africanized Apis mellifera (Hymenoptera: Apidae). Microscopy Research and Technique, 77(4):274–281, 2014. ISSN 1097-0029. doi: 10.1002/jemt.22339. URL https://onlinelibrary. wiley.com/doi/abs/10.1002/jemt.22339.

[132] Karen-Anne Neufeld and Jane A. Foster. Effects of gut microbiota on the brain: implications for psychiatry. Journal of Psychiatry & Neuroscience : JPN, 34(3): 230–231, May 2009. ISSN 1180-4882. URL https://www.ncbi.nlm.nih.gov/ pmc/articles/PMC2674977/.

296 [133] Gloria DeGrandi-Hoffman, Vanessa Corby-Harris, Emily Watkins DeJong, Mona Chambers, and Geoffrey Hidalgo. Honey bee gut microbial communities are robust to the fungicide Pristine® consumed in pollen. Apidologie, 48(3):340–352, May 2017. ISSN 0044-8435, 1297-9678. doi: 10.1007/s13592-016-0478-y. URL https://link.springer.com/article/10.1007/s13592-016-0478-y.

[134] V. Girolami, L. Mazzon, A. Squartini, N. Mori, M. Marzaro, A. Di bernardo, M. Greatti, C. Giorio, and A. Tapparo. Translocation of Neonicotinoid Insecti- cides From Coated Seeds to Seedling Guttation Drops: A Novel Way of Intoxica- tion for Bees. Journal of Economic Entomology, 102(5):1808–1815, October 2009. ISSN 0022-0493. doi: 10.1603/029.102.0511. URL https://academic.oup.com/ jee/article/102/5/1808/2199218.

[135] David Guez, Séverine Suchail, Monique Gauthier, Ryszard Maleszka, and Luc P. Belzunces. Contrasting Effects of Imidacloprid on Habituation in 7- and 8- Day-Old Honeybees (Apis mellifera). Neurobiology of Learning and Mem- ory, 76(2):183–191, September 2001. ISSN 1074-7427. doi: 10.1006/nlme. 2000.3995. URL http://www.sciencedirect.com/science/article/pii/ S1074742700939959.

[136] Axel Decourtye, Eric Lacassie, and Minh-Hà Pham-Delègue. Learning perfor- mances of honeybees (Apis mellifera L) are differentially affected by imidacloprid according to the season. Pest Management Science, 59(3):269–278, 2003. ISSN 1526-4998. doi: 10.1002/ps.631. URL https://onlinelibrary.wiley. com/doi/abs/10.1002/ps.631.

[137] Julia C Jones, Carmelo Fruciano, Falk Hildebrand, Hasan Al Toufalilia, Nicholas J Balfour, Peer Bork, Philipp Engel, Francis LW Ratnieks, and William OH Hughes. Gut microbiota composition is associated with environmental landscape in honey bees. Ecology and Evolution, 8(1):441–451, November 2017. ISSN 2045- 7758. doi: 10.1002/ece3.3597. URL https://www.ncbi.nlm.nih.gov/pmc/ articles/PMC5756847/.

297 [138] Michael. Environment Secretary backs further restrictions on neonicotinoid pesticides, 2017. URL https://www.gov.uk/government/news/ environment-secretary-backs-further-restrictions-on-neonicotinoid-pesticides.

[139] Aaron D. Goldberg, C. David Allis, and Emily Bernstein. Epigenetics: A Land- scape Takes Shape. Cell, 128(4):635–638, February 2007. ISSN 0092-8674. doi: 10.1016/j.cell.2007.02.006. URL http://www.sciencedirect.com/science/ article/pii/S0092867407001869.

[140] M. Collotta, P. A. Bertazzi, and V. Bollati. Epigenetics and pesticides. Toxicology, 307:35–41, May 2013. ISSN 0300-483X. doi: 10.1016/j.tox. 2013.01.017. URL http://www.sciencedirect.com/science/article/pii/ S0300483X1300022X.

[141] Hua Yan, Roberto Bonasio, Daniel F.Simola, Jürgen Liebig, Shelley L. Berger, and Danny Reinberg. DNA Methylation in Social Insects: How Epigenetics Can Con- trol Behavior and Longevity. Annual Review of Entomology, 60(1):435–452, 2015. doi: 10.1146/annurev-ento-010814-020803. URL https://doi.org/10.1146/ annurev-ento-010814-020803.

[142] R. Satta, E. Maloku, A. Zhubi, F.Pibiri, M. Hajos, E. Costa, and A. Guidotti. Nico- tine decreases DNA methyltransferase 1 expression and glutamic acid decarboxylase 67 promoter methylation in GABAergic interneurons. Proceedings of the National Academy of Sciences, October 2008. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.0808699105. URL http://www.pnas.org/content/early/ 2008/10/10/0808699105.

[143] Yasufumi Shutoh, Makio Takeda, Ryoichi Ohtsuka, Atsuko Haishima, Satoru Ya- maguchi, Hideaki Fujie, Yutaka Komatsu, Keizo Maita, and Takanori Harada. Low dose effects of dichlorodiphenyltrichloroethane (DDT) on gene transcription and DNA methylation in the hypothalamus of young male rats: implication of hormesis-like effects. The Journal of Toxicological Sciences, 34(5): 469–482, October 2009. ISSN 0388-1350, 1880-3989. doi: 10.2131/jts.34. 298 469. URL https://www.jstage.jst.go.jp/article/jts/34/5/34_5_469/ _article/-char/ja/.

[144] A. Oppold, A. Kreß, J. Vanden Bussche, J. B. Diogo, U. Kuch, J. Oehlmann, M. B. Vandegehuchte, and R. Müller. Epigenetic alterations and decreasing insecticide sensitivity of the Asian tiger mosquito Aedes albopictus. Ecotoxicology and Environmental Safety, 122:45–53, December 2015. ISSN 0147-6513. doi: 10. 1016/j.ecoenv.2015.06.036. URL http://www.sciencedirect.com/science/ article/pii/S0147651315300063.

[145] Daniel Desaulniers, Gong-hua Xiao, Hong Lian, Yong-Lai Feng, Jiping Zhu, Jamie Nakai, and Wayne J. Bowers. Effects of Mixtures of Polychlorinated Biphenyls, Methylmercury, and Organochlorine Pesticides on Hepatic DNA Methylation in Prepubertal Female Sprague-Dawley Rats. International Journal of Toxicology, 28(4):294–307, July 2009. ISSN 1091-5818. doi: 10.1177/1091581809337918. URL https://doi.org/10.1177/1091581809337918.

[146] Liangfeng Han, P. Dane Witmer, Emily Casey, David Valle, and Saraswati Suku- mar. DNA methylation regulates MicroRNA expression. Cancer Biology & Ther- apy, 6(8):1284–1288, August 2007. ISSN 1555-8576.

[147] Karl M. Glastad, Brendan G. Hunt, Soojin V. Yi, and Michael A. D. Goodisman. Epigenetic inheritance and genome regulation: is DNA methylation linked to ploidy in haplodiploid insects? Proceedings of the Royal Society B: Biological Sciences, 281(1785), June 2014. ISSN 0962-8452. doi: 10.1098/rspb.2014.0411. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4024298/.

[148] Kin Kuan Hoi, Phillip J. Daborn, Paul Battlay, Charles Robin, Philip Batterham, Richard A. J. O’Hair, and William A. Donald. Dissecting the insect metabolic ma- chinery using twin ion mass spectrometry: a single P450 enzyme metabolizing the insecticide imidacloprid in vivo. Analytical Chemistry, 86(7):3525–3532, April 2014. ISSN 1520-6882. doi: 10.1021/ac404188g.

299 [149] Iris Karunker, Juergen Benting, Bettina Lueke, Tanja Ponge, Ralf Nauen, Em- manouil Roditakis, John Vontas, Kevin Gorman, Ian Denholm, and Shai Morin. Over-expression of cytochrome P450 CYP6cm1 is associated with high resistance to imidacloprid in the B and Q biotypes of Bemisia tabaci (Hemiptera: Aleyrodidae). Insect Biochemistry and Molecular Biology, 38(6):634–644, June 2008. ISSN 0965-1748. doi: 10.1016/j.ibmb.2008.03.008. URL http://www. sciencedirect.com/science/article/pii/S096517480800060X.

[150] Richard H. ffrench Constant. The Molecular Genetics of Insecticide Resis- tance. Genetics, 194(4):807–815, August 2013. ISSN 0016-6731. doi: 10.1534/ genetics.112.141895. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC3730913/.

[151] Oliver Bossdorf, Christina L. Richards, and Massimo Pigliucci. Epigenetics for ecologists. Ecology Letters, 11(2):106–115, February 2008. ISSN 1461-0248. doi: 10.1111/j.1461-0248.2007.01130.x. URL https://onlinelibrary.wiley.com/ doi/abs/10.1111/j.1461-0248.2007.01130.x.

[152] Jessica A. Head. Patterns of DNA Methylation in Animals: An Ecotoxicological Perspective. Integrative and Comparative Biology, 54(1):77–86, July 2014. ISSN 1540-7063. doi: 10.1093/icb/icu025. URL https://academic.oup.com/icb/ article/54/1/77/2797781.

[153] Michiel B. Vandegehuchte and Colin R. Janssen. Epigenetics in an ecotoxicological context. Mutation Research/Genetic Toxicology and Environmental Mu- tagenesis, 764-765:36–45, April 2014. ISSN 1383-5718. doi: 10.1016/j.mrgentox. 2013.08.008. URL http://www.sciencedirect.com/science/article/pii/ S1383571813002325.

[154] Adam J. Bewick, Kevin J. Vogel, Allen J. Moore, and Robert J. Schmitz. Evolution of DNA Methylation across Insects. Molecular Biology and Evolution, 34(3):654– 665, March 2017. ISSN 0737-4038. doi: 10.1093/molbev/msw264. URL https: //academic.oup.com/mbe/article/34/3/654/2739700. 300 [155] Laura Welsh, Ryszard Maleszka, and Sylvain Foret. Detecting rare asym- metrically methylated cytosines and decoding methylation patterns in the honeybee genome. Royal Society Open Science, 4(9):170248, September 2017. ISSN 2054-5703. doi: 10.1098/rsos.170248. URL http://rsos. royalsocietypublishing.org/content/4/9/170248.

[156] Sandra M. Rehan, Karl M. Glastad, Sarah P. Lawson, and Brendan G. Hunt. The Genome and Methylome of a Subsocial Small Carpenter Bee, Ceratina calcarata. Genome Biology and Evolution, 8(5):1401–1410, May 2016. doi: 10.1093/gbe/ evw079. URL https://academic.oup.com/gbe/article/8/5/1401/2939574.

[157] Suhua Feng, Shawn J. Cokus, Xiaoyu Zhang, Pao-Yang Chen, Magnolia Bostick, Mary G. Goll, Jonathan Hetzel, Jayati Jain, Steven H. Strauss, Marnie E. Halpern, Chinweike Ukomadu, Kirsten C. Sadler, Sriharsa Pradhan, Matteo Pellegrini, and Steven E. Jacobsen. Conservation and divergence of methylation patterning in plants and animals. Proceedings of the National Academy of Sciences, 107(19):8689–8694, May 2010. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas. 1002720107. URL http://www.pnas.org/content/107/19/8689.

[158] Aneta Strachecka, Jacek Chobotow, Jerzy Paleolog, Aleksandra Ło´s, Michał Schulz, Dariusz Teper, Halina Kucharczyk, and Maciej Grzybek. Insights into the biochemical defence and methylation of the solitary bee Osmia rufa L: A foundation for examining eusociality development. PLOS ONE, 12(4): e0176539, April 2017. ISSN 1932-6203. doi: 10.1371/journal.pone.0176539. URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0176539.

[159] Mary Grace Goll and Timothy H. Bestor. Eukaryotic Cytosine Methyltrans- ferases. Annual Review of Biochemistry, 74(1):481–514, 2005. doi: 10. 1146/annurev.biochem.74.010904.153721. URL https://doi.org/10.1146/ annurev.biochem.74.010904.153721.

301 [160] Sylvain Foret, Robert Kucharski, Matteo Pellegrini, Suhua Feng, Steven E. Ja- cobsen, Gene E. Robinson, and Ryszard Maleszka. DNA methylation dynamics, metabolic ﬂuxes, gene splicing, and alternative phenotypes in honey bees. Pro- ceedings of the National Academy of Sciences, 109(13):4968–4973, March 2012. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.1202392109. URL https://www. pnas.org/content/109/13/4968.

[161] Harindra E. Amarasinghe, Crisenthiya I. Clayton, and Eamonn B. Mallon. Methy- lation and worker reproduction in the bumble-bee (Bombus terrestris). Proc. R. Soc. B, 281(1780):20132502, April 2014. ISSN 0962-8452, 1471-2954. doi: 10.1098/rspb.2013.2502. URL http://rspb.royalsocietypublishing.org/ content/281/1780/20132502.

[162] Hongmei Li-Byarlay. The Function of DNA Methylation Marks in Social Insects. Frontiers in Ecology and Evolution, 4, 2016. ISSN 2296-701X. doi: 10.3389/fevo. 2016.00057. URL https://www.frontiersin.org/articles/10.3389/fevo. 2016.00057/full#B60.

[163] Hua Yan, Daniel F.Simola, Roberto Bonasio, Jürgen Liebig, Shelley L. Berger, and Danny Reinberg. Eusocial insects as emerging models for behavioural epigenetics. Nature Reviews Genetics, 15(10):677–688, October 2014. ISSN 1471-0064. doi: 10.1038/nrg3787. URL https://www.nature.com/articles/nrg3787.

[164] En Li and Yi Zhang. DNA Methylation in Mammals. Cold Spring Har- bor Perspectives in Biology, 6(5), May 2014. ISSN 1943-0264. doi: 10.1101/ cshperspect.a019133. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC3996472/.

[165] Ying Wang, Mireia Jorda, Peter L. Jones, Ryszard Maleszka, Xu Ling, Hugh M. Robertson, Craig A. Mizzen, Miguel A. Peinado, and Gene E. Robinson. Func- tional CpG methylation system in a social insect. Science (New York, N.Y.), 314 (5799):645–647, October 2006. ISSN 1095-9203. doi: 10.1126/science.1135213.

302 [166] Miho M. Suzuki, Alastair R. W. Kerr, Dina De Sousa, and Adrian Bird. CpG methylation is targeted to transcription units in an invertebrate genome. Genome Re- search, 17(5):625–631, May 2007. ISSN 1088-9051. doi: 10.1101/gr.6163007.

[167] Jia Zeng and Soojin V.Yi. DNA Methylation and Genome Evolution in Honeybee: Gene Length, Expression, Functional Enrichment Covary with the Evolutionary Signature of DNA Methylation. Genome Biology and Evolution, 2:770–780, 2010. ISSN 1759-6653. doi: 10.1093/gbe/evq060. URL https://www.ncbi.nlm.nih. gov/pmc/articles/PMC2975444/.

[168] Navin Elango, Brendan G. Hunt, Michael A. D. Goodisman, and Soojin V.Yi. DNA methylation is widespread and associated with differential gene expression in castes of the honeybee, Apis mellifera. Proceedings of the National Academy of Sciences of the United States of America, 106(27):11206–11211, July 2009. ISSN 0027-8424. doi: 10.1073/pnas.0900301106. URL https://www.ncbi.nlm.nih. gov/pmc/articles/PMC2708677/.

[169] Auinash Kalsotra and Thomas A. Cooper. Functional consequences of developmentally regulated alternative splicing. Nature Reviews. Genetics, 12(10):715– 729, September 2011. ISSN 1471-0056. doi: 10.1038/nrg3052. URL https: //www.ncbi.nlm.nih.gov/pmc/articles/PMC3321218/.

[170] Samuel V. Arsenault, Brendan G. Hunt, and Sandra M. Rehan. The effect of maternal care on gene expression and DNA methylation in a subsocial bee. Nature Communications, 9(1):3468, August 2018. ISSN 2041-1723. doi: 10.1038/s41467-018-05903-0. URL https://www.nature.com/articles/ s41467-018-05903-0.

[171] Veeranan Chaimanee, Jay D. Evans, Yanping Chen, Caitlin Jackson, and Jeffery S. Pettis. Sperm viability and gene expression in honey bee queens (Apis mellifera) following exposure to the neonicotinoid insecticide imidacloprid and the organophosphate acaricide coumaphos. Journal of Insect Physiology, 89:1–8,

303 June 2016. ISSN 0022-1910. doi: 10.1016/j.jinsphys.2016.03.004. URL http: //www.sciencedirect.com/science/article/pii/S0022191016300270.

[172] Kamila Derecka, Martin J. Blythe, Sunir Malla, Diane P. Genereux, Alessandro Guffanti, Paolo Pavan, Anna Moles, Charles Snart, Thomas Ryder, Catharine A. Ortori, David A. Barrett, Eugene Schuster, and Reinhard Stöger. Transient Expo- sure to Low Levels of Insecticide Affects Metabolic Networks of Honeybee Larvae. PLOS ONE, 8(7):e68191, July 2013. ISSN 1932-6203. doi: 10.1371/journal.pone. 0068191. URL http://journals.plos.org/plosone/article?id=10.1371/ journal.pone.0068191.

[173] Julie Aufauvre, Barbara Misme-Aucouturier, Bernard Viguès, Catherine Tex- ier, Frédéric Delbac, and Nicolas Blot. Transcriptome Analyses of the Hon- eybee Response to Nosema ceranae and Insecticides. PLOS ONE, 9(3): e91686, March 2014. ISSN 1932-6203. doi: 10.1371/journal.pone.0091686. URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0091686.

[174] Jaime M. Tovar-Corona, Atahualpa Castillo-Morales, Lu Chen, Brett P. Olds, John M. Clark, Stuart E. Reynolds, Barry R. Pittendrigh, Edward J. Feil, and Araxi O. Urrutia. Alternative Splice in Alternative Lice. Molecular Biology and Evolution, 32(10):2749–2759, October 2015. ISSN 0737-4038. doi: 10. 1093/molbev/msv151. URL https://academic.oup.com/mbe/article/32/ 10/2749/1212152.

[175] Mark C. Harrison, Robert L. Hammond, and Eamonn B. Mallon. Reproduc- tive workers show queenlike gene expression in an intermediately eusocial insect, the buff-tailed bumble bee Bombus terrestris. Molecular Ecology, 24(12): 3043–3063, 2015. ISSN 1365-294X. doi: 10.1111/mec.13215. URL https: //onlinelibrary.wiley.com/doi/abs/10.1111/mec.13215.

304 [176] Simon Andrews. Babraham Bioinformatics - FastQC A Quality Control tool for High Throughput Sequence Data. URL https://www.bioinformatics. babraham.ac.uk/projects/fastqc/.

[177] Marcel Martin. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal, 17(1):pp. 10–12, May 2011. ISSN 2226-6089. URL http://journal.embnet.org/index.php/embnetjournal/ article/view/200.

[178] Anthony M. Bolger, Marc Lohse, and Bjoern Usadel. Trimmomatic: a ﬂexi- ble trimmer for Illumina sequence data. Bioinformatics, 30(15):2114–2120, Au- gust 2014. ISSN 1367-4803. doi: 10.1093/bioinformatics/btu170. URL https: //academic.oup.com/bioinformatics/article/30/15/2114/2390096.

[179] Ben M. Sadd, Seth M. Barribeau, Guy Bloch, Dirk C. de Graaf, Peter Dearden, Christine G. Elsik, Jürgen Gadau, Cornelis JP Grimmelikhuijzen, Martin Has- selmann, Jeffrey D. Lozier, Hugh M. Robertson, Guy Smagghe, Eckart Stolle, Matthias Van Vaerenbergh, Robert M. Waterhouse, Erich Bornberg-Bauer, Stef- fen Klasberg, Anna K. Bennett, Francisco Câmara, Roderic Guigó, Katharina Hoff, Marco Mariotti, Monica Munoz-Torres, Terence Murphy, Didac Santes- masses, Gro V. Amdam, Matthew Beckers, Martin Beye, Matthias Biewer, Már- cia MG Bitondi, Mark L. Blaxter, Andrew FG Bourke, Mark JF Brown, Sever- ine D. Buechel, Rossanah Cameron, Kaat Cappelle, James C. Carolan, Olivier Christiaens, Kate L. Ciborowski, David F. Clarke, Thomas J. Colgan, David H. Collins, Andrew G. Cridge, Tamas Dalmay, Stephanie Dreier, Louis du Plessis, Elizabeth Duncan, Silvio Erler, Jay Evans, Tiago Falcon, Kevin Flores, Flávia CP Freitas, Taro Fuchikawa, Tanja Gempe, Klaus Hartfelder, Frank Hauser, So- phie Helbing, Fernanda C. Humann, Frano Irvine, Lars S. Jermiin, Claire E. Johnson, Reed M. Johnson, Andrew K. Jones, Tatsuhiko Kadowaki, Jonathan H. Kidner, Vasco Koch, Arian Köhler, F. Bernhard Kraus, H. Michael G. Lattorff,

305 Megan Leask, Gabrielle A. Lockett, Eamonn B. Mallon, David S. Marco Anto- nio, Monika Marxer, Ivan Meeus, Robin FA Moritz, Ajay Nair, Kathrin Näpﬂin, Inga Nissen, Jinzhi Niu, Francis MF Nunes, John G. Oakeshott, Amy Osborne, Marianne Otte, Daniel G. Pinheiro, Nina Rossié, Olav Rueppell, Carolina G. Santos, Regula Schmid-Hempel, Björn D. Schmitt, Christina Schulte, Zilá LP Simões, Michelle PM Soares, Luc Swevers, Eva C. Winnebeck, Florian Wolschin, Na Yu, Evgeny M. Zdobnov, Peshtewani K. Aqrawi, Kerstin P.Blankenburg, Mar- cus Coyle, Liezl Francisco, Alvaro G. Hernandez, Michael Holder, Matthew E. Hudson, LaRonda Jackson, Joy Jayaseelan, Vandita Joshi, Christie Kovar, San- dra L. Lee, Robert Mata, Tittu Mathew, Irene F. Newsham, Robin Ngo, Geof- frey Okwuonu, Christopher Pham, Ling-Ling Pu, Nehad Saada, Jireh Santibanez, DeNard Simmons, Rebecca Thornton, Aarti Venkat, Kimberly KO Walden, Yuan- Qing Wu, Griet Debyser, Bart Devreese, Claire Asher, Julie Blommaert, Ariel D. Chipman, Lars Chittka, Bertrand Fouks, Jisheng Liu, Meaghan P.O’Neill, Seirian Sumner, Daniela Puiu, Jiaxin Qu, Steven L. Salzberg, Steven E. Scherer, Donna M. Muzny, Stephen Richards, Gene E. Robinson, Richard A. Gibbs, Paul Schmid- Hempel, and Kim C. Worley. The genomes of two key bumblebee species with primitive eusocial organization. Genome Biology, 16:76, April 2015. ISSN 1465-6906. doi: 10.1186/s13059-015-0623-3. URL https://doi.org/10.1186/ s13059-015-0623-3.

[180] Felix Krueger and Simon R. Andrews. Bismark: a ﬂexible aligner and methylation caller for Bisulﬁte-Seq applications. Bioinformatics, 27(11):1571–1572, June 2011. ISSN 1367-4803. doi: 10.1093/bioinformatics/btr167. URL https: //academic.oup.com/bioinformatics/article/27/11/1571/216956.

[181] Ben Langmead and Steven L Salzberg. Fast gapped-read alignment with Bowtie 2. Nature methods, 9(4):357–359, March 2012. ISSN 1548-7091. doi: 10.1038/nmeth.1923. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC3322381/.

306 [182] G rothendieck. sqldf: Manipulate R Data Frames Using SQL version 0.4-11 from CRAN, 2017. URL https://rdrr.io/cran/sqldf/.

[183] R Development Core Team. R: The R Project for Statistical Computing, 2008. URL https://www.r-project.org/.

[184] Altuna Akalin, Matthias Kormaksson, Sheng Li, Francine E. Garrett-Bakelman, Maria E. Figueroa, Ari Melnick, and Christopher E. Mason. methylKit: a com- prehensive R package for the analysis of genome-wide DNA methylation pro- ﬁles. Genome Biology, 13(10):R87, October 2012. ISSN 1474-760X. doi: 10.1186/ gb-2012-13-10-r87.

[185] Michael Lawrence, Wolfgang Huber, Hervé Pagès, Patrick Aboyoun, Marc Carl- son, Robert Gentleman, Martin T. Morgan, and Vincent J. Carey. Software for Computing and Annotating Genomic Ranges. PLOS Computational Biol- ogy, 9(8):e1003118, August 2013. ISSN 1553-7358. doi: 10.1371/journal.pcbi. 1003118. URL http://journals.plos.org/ploscompbiol/article?id=10. 1371/journal.pcbi.1003118.

[186] Martin Morgan, Herv\’e Pag\‘es, Valerie Obenchain, and Nathaniel Hayden. Rsamtools: Binary alignment (BAM), FASTA, variant call (BCF), and tabix ﬁle import version 1.32.0 from Bioconductor. URL http://bioconductor.org/ packages/release/bioc/html/Rsamtools.html.

[187] Simon Anders, Alejandro Reyes, and Wolfgang Huber. Detecting differential usage of exons from RNA-seq data. Genome Research, 22(10):2008–2017, October 2012. ISSN 1088-9051. doi: 10.1101/gr.133744.111. URL https://www.ncbi. nlm.nih.gov/pmc/articles/PMC3460195/.

[188] Michael I. Love, Wolfgang Huber, and Simon Anders. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology, 15: 550, December 2014. ISSN 1474-760X. doi: 10.1186/s13059-014-0550-8. URL https://doi.org/10.1186/s13059-014-0550-8.

307 [189] Daehwan Kim, Ben Langmead, and Steven L. Salzberg. HISAT: a fast spliced aligner with low memory requirements. Nature Methods, 12(4):357–360, April 2015. ISSN 1548-7105. doi: 10.1038/nmeth.3317. URL https://www.nature. com/articles/nmeth.3317.

[190] Heng Li, Bob Handsaker, Alec Wysoker, Tim Fennell, Jue Ruan, Nils Homer, Ga- bor Marth, Goncalo Abecasis, Richard Durbin, and 1000 Genome Project Data Processing Subgroup. The Sequence Alignment/Map format and SAMtools. Bioinformatics (Oxford, England), 25(16):2078–2079, August 2009. ISSN 1367- 4811. doi: 10.1093/bioinformatics/btp352.

[191] Mihaela Pertea, Geo M. Pertea, Corina M. Antonescu, Tsung-Cheng Chang, Joshua T. Mendell, and Steven L. Salzberg. StringTie enables improved recon- struction of a transcriptome from RNA-seq reads. Nature Biotechnology, 33(3): 290–295, March 2015. ISSN 1546-1696. doi: 10.1038/nbt.3122. URL https: //www.nature.com/articles/nbt.3122.

[192] Charlotte Rich, Marco Reitz, Ruth Eichmann, Sophie Hermann, Dafyd J. Jenk- ins, Karl-Heinz Kogel, Eddi Esteban, Sascha Ott, and Patrick Schafer. Cell type identity determines transcriptomic immune responses in Arabidopsis thaliana roots. bioRxiv, page 302448, May 2018. doi: 10.1101/302448. URL https: //www.biorxiv.org/content/early/2018/05/23/302448.

[193] Yoav Benjamini and Yosef Hochberg. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Sta- tistical Society. Series B (Methodological), 57(1):289–300, 1995. ISSN 0035-9246. URL http://www.jstor.org/stable/2346101.

[194] Alyssa C Frazee, Geo Pertea, Andrew E Jaffe, Ben Langmead, Steven L Salzberg, and Jeffrey T Leek. Ballgown bridges the gap between transcriptome assembly and expression analysis. Nature biotechnology, 33(3):243–246, March 2015. ISSN 1087-0156. doi: 10.1038/nbt.3172. URL https://www.ncbi.nlm.nih. gov/pmc/articles/PMC4792117/. 308 [195] John D. Storey and Robert Tibshirani. Statistical signiﬁcance for genomewide studies. Proceedings of the National Academy of Sciences, 100(16):9440–9445, Au- gust 2003. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.1530509100. URL http://www.pnas.org/content/100/16/9440.

[196] François Olivier Hébert, Stephan Grambauer, Iain Barber, Christian R. Landry, and Nadia Aubin-Horth. Transcriptome sequences spanning key developmental states as a resource for the study of the cestode Schistocephalus solidus, a threespine stickleback parasite. GigaScience, 5:24, June 2016. ISSN 2047- 217X. doi: 10.1186/s13742-016-0128-3. URL https://doi.org/10.1186/ s13742-016-0128-3.

[197] Ana Conesa, Stefan Götz, Juan Miguel García-Gómez, Javier Terol, Manuel Talón, and Montserrat Robles. Blast2go: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics (Oxford, England), 21(18):3674–3676, September 2005. ISSN 1367-4803. doi: 10.1093/ bioinformatics/bti610.

[198] David Amar, Itziar Frades, Agnieszka Danek, Tatyana Goldberg, Sanjeev K. Sharma, Pete E. Hedley, Estelle Proux-Wera, Erik Andreasson, Ron Shamir, Oren Tzfadia, and Erik Alexandersson. Evaluation and integration of functional annotation pipelines for newly sequenced organisms: the potato genome as a test case. BMC Plant Biology, 14:329, December 2014. ISSN 1471-2229. doi: 10.1186/ s12870-014-0329-9. URL https://doi.org/10.1186/s12870-014-0329-9.

[199] S. Falcon and R. Gentleman. Using GOstats to test gene lists for GO term association. Bioinformatics, 23(2):257–258, January 2007. ISSN 1367-4803. doi: 10.1093/ bioinformatics/btl567. URL https://academic.oup.com/bioinformatics/ article/23/2/257/204776.

[200] Fran Supek, Matko Bošnjak, Nives Škunca, and Tomislav Šmuc. REVIGO Sum- marizes and Visualizes Long Lists of Gene Ontology Terms. PLOS ONE, 6 (7):e21800, July 2011. ISSN 1932-6203. doi: 10.1371/journal.pone.0021800. 309 URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0021800.

[201] Guangchuang Yu, Li-Gen Wang, Yanyan Han, and Qing-Yu He. clusterProﬁler: an R package for comparing biological themes among gene clusters. Omics: A Journal of Integrative Biology, 16(5):284–287, May 2012. ISSN 1557-8100. doi: 10.1089/omi.2011.0118.

[202] Jérôme Teulière, Christelle Gally, Gian Garriga, Michel Labouesse, and Elisabeth Georges-Labouesse. MIG-15 and ERM-1 promote growth cone directional migration in parallel to UNC-116 and WVE-1. Development, 138(20):4475–4485, October 2011. ISSN 0950-1991, 1477-9129. doi: 10.1242/dev.061952. URL http://dev.biologists.org/content/138/20/4475.

[203] Lihua Zheng, Yehudit Michelson, Vita Freger, Ziva Avraham, Koen J. T. Venken, Hugo J. Bellen, Monica J. Justice, and Ron Wides. Drosophila Ten-m and Filamin Affect Motor Neuron Growth Cone Guidance. PLOS ONE, 6(8): e22956, August 2011. ISSN 1932-6203. doi: 10.1371/journal.pone.0022956. URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0022956.

[204] Maria Del Carmen Laﬁta-Navarro, Rosa Blanco, Jorge Mata-Garrido, Judit Liaño- Pons, Olga Tapia, Lucía García-Gutiérrez, Eva García-Alegría, María T. Berciano, Miguel Lafarga, and Javier León. MXD1 localizes in the nucleolus, binds UBF and impairs rRNA synthesis. Oncotarget, 7(43):69536–69548, 2016. ISSN 1949-2553. doi: 10.18632/oncotarget.11766.

[205] Fabiana M. Meliso, Danilo Micali, Camila T. Silva, Thaís S. Sabedot, Simon G. Co- etzee, Adrian Koch, Fabian B. Fahlbusch, Houtan Noushmehr, Regine Schneider- Stock, and Miriam G. Jasiulionis. SIRT1 regulates Mxd1 during malignant melanoma progression. Oncotarget, 8(70):114540–114553, October 2017. ISSN 1949-2553. doi: 10.18632/oncotarget.21457. URL https://www.ncbi.nlm. nih.gov/pmc/articles/PMC5777712/. 310 [206] Cynthia A. Bayer, Susan R. Halsell, James W. Fristrom, Daniel P.Kiehart, and Lau- rence von Kalm. Genetic Interactions Between the RhoA and Stubble-stubbloid Loci Suggest a Role for a Type II Transmembrane Serine Protease in Intracel- lular Signaling During Drosophila Imaginal Disc Morphogenesis. Genetics, 165 (3):1417–1432, November 2003. ISSN 0016-6731, 1943-2631. URL http://www. genetics.org/content/165/3/1417.

[207] René Eickhoff and Gerd Bicker. Developmental expression of cell recognition molecules in the mushroom body and antennal lobe of the locust Locusta migratoria. The Journal of Comparative Neurology, 520(9):2021–2040, June 2012. ISSN 1096-9861. doi: 10.1002/cne.23026.

[208] Nicole Berndt, Ekaterina Seib, Soya Kim, Tobias Troost, Marvin Lyga, Jessica Lan- genbach, Sebastian Haensch, Konstantina Kalodimou, Christos Delidakis, and Thomas Klein. Ubiquitylation-independent activation of Notch signalling by Delta. eLife, 6, 2017. ISSN 2050-084X. doi: 10.7554/eLife.27346.

[209] Kyucheol Cho, Travis G. Vaught, Hong Ji, Dongmin Gu, Catherine Papasakelariou-Yared, Nicola Horstmann, Jean Marie Jennings, Moonsup Lee, Lisa M. Sevilla, Malgorzata Kloc, Albert B. Reynolds, Fiona M. Watt, Richard G. Brennan, Andrew P. Kowalczyk, and Pierre D. McCrea. Xenopus Kazrin in- teracts with ARVCF-catenin, spectrin and p190b RhoGAP, and modulates RhoA activity and epithelial integrity. Journal of Cell Science, page jcs.072041, January 2010. ISSN 0021-9533, 1477-9137. doi: 10.1242/jcs.072041. URL http://jcs.biologists.org/content/early/2010/11/09/jcs.072041.

[210] D. Yamamoto. Signaling mechanisms in induction of the R7 photoreceptor in the developing Drosophila retina. BioEssays: News and Reviews in Molecular, Cellular and Developmental Biology, 16(4):237–244, April 1994. ISSN 0265-9247. doi: 10.1002/bies.950160406.

311 [211] Tadahiro Goda, Xin Tang, Yujiro Umezaki, Michelle L. Chu, Michael Kunst, Michael N. Nitabach, and Fumika N. Hamada. Drosophila DH31 Neuropep- tide and PDF Receptor Regulate Night-Onset Temperature Preference. Journal of Neuroscience, 36(46):11739–11754, November 2016. ISSN 0270-6474, 1529- 2401. doi: 10.1523/JNEUROSCI.0964-16.2016. URL http://www.jneurosci. org/content/36/46/11739.

[212] Jesús Romero-Pozuelo, Constantinos Demetriades, Phillip Schroeder, and Aure- lio A. Teleman. CycD/Cdk4 and Discontinuities in Dpp Signaling Activate TORC1 in the Drosophila Wing Disc. Developmental Cell, 42(4):376–387.e5, 2017. ISSN 1878-1551. doi: 10.1016/j.devcel.2017.07.019.

[213] Emanuele Brai, Swananda Marathe, Lorena Zentilin, Mauro Giacca, Johannes Nimpf, Robert Kretz, Alessandra Scotti, and Lavinia Alberi. Notch1 activity in the olfactory bulb is odour-dependent and contributes to olfactory behaviour. The European Journal of Neuroscience, 40(10):3436–3449, November 2014. ISSN 1460-9568. doi: 10.1111/ejn.12719.

[214] Lavinia Alberi, Shuxi Liu, Yue Wang, Ramy Badie, Constance Smith-Hicks, Jing Wu, Tarran J. Pierfelice, Bagrat Abazyan, Mark P.Mattson, Dietmar Kuhl, Mikhail Pletnikov, Paul F. Worley, and Nicholas Gaiano. Activity-Induced Notch Signal- ing in Neurons Requires Arc/Arg3.1 and Is Essential for Synaptic Plasticity in Hippocampal Networks. Neuron, 69(3):437–444, February 2011. ISSN 0896- 6273. doi: 10.1016/j.neuron.2011.01.004. URL http://www.sciencedirect. com/science/article/pii/S0896627311000298.

[215] Cole Trapnell, David G. Hendrickson, Martin Sauvageau, Loyal Goff, John L. Rinn, and Lior Pachter. Differential analysis of gene regulation at transcript resolution with RNA-seq. Nature Biotechnology, 31(1):46–53, January 2013. ISSN 1546-1696. doi: 10.1038/nbt.2450. URL https://www.nature.com/articles/ nbt.2450.

312 [216] Tetsuo Yasugi, Takayuki Yamada, and Takashi Nishimura. Adaptation to dietary conditions by trehalose metabolism in Drosophila. Scientiﬁc Reports, 7, May 2017. ISSN 2045-2322. doi: 10.1038/s41598-017-01754-9. URL https: //www.ncbi.nlm.nih.gov/pmc/articles/PMC5431645/.

[217] Benchamart Moolmuang and Michael A. Tainsky. CREG1 enhances p16(INK4a) - induced cellular senescence. Cell Cycle (Georgetown, Tex.), 10(3):518–530, Febru- ary 2011. ISSN 1551-4005. doi: 10.4161/cc.10.3.14756.

[218] Ágnes Bánréti, Tamás Lukácsovich, György Csikós, Miklós Erdélyi, and Miklós Sass. PP2a regulates autophagy in two alternative ways in Drosophila. Au- tophagy, 8(4):623–636, April 2012. ISSN 1554-8627. doi: 10.4161/auto.19081. URL https://doi.org/10.4161/auto.19081.

[219] Ana Terriente-Félix, Ana López-Varea, and Jose F. de Celis. Identiﬁcation of Genes Affecting Wing Patterning Through a Loss-of-Function Mutagenesis Screen and Characterization of med15 Function During Wing Development. Ge- netics, 185(2):671–684, June 2010. ISSN 0016-6731. doi: 10.1534/genetics.109. 113670. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2881146/.

[220] Beth E. Stronach and Norbert Perrimon. Stress signaling in drosophila. Onco- gene, 18(45):6172–6182, November 1999. ISSN 1476-5594. doi: 10.1038/sj.onc. 1203125. URL https://www.nature.com/articles/1203125.

[221] Eva C. Winnebeck, Craig D. Millar, and Guy R. Warman. Why Does Insect RNA Look Degraded? Journal of Insect Science, 10, September 2010. ISSN 1536- 2442. doi: 10.1673/031.010.14119. URL https://www.ncbi.nlm.nih.gov/ pmc/articles/PMC3016993/.

[222] Terrence RJ Lappin, David G Grier, Alexander Thompson, and Henry L Halliday. HOX GENES: Seductive Science, Mysterious Mechanisms. The Ulster Medical Journal, 75(1):23–31, January 2006. ISSN 0041-6193. URL https://www.ncbi. nlm.nih.gov/pmc/articles/PMC1891803/.

313 [223] Chenghao Chen, Edgar Buhl, Min Xu, Vincent Croset, Johanna S. Rees, Kathryn S. Lilley, Richard Benton, James J. L. Hodge, and Ralf Stanewsky. Drosophila Ionotropic Receptor 25a mediates circadian clock resetting by temperature. Nature, 527(7579):516–520, November 2015. ISSN 1476-4687. doi: 10.1038/ nature16148.

[224] Liudy G. Hernández, Bingwen Lu, Gabriel C. N. da Cruz, Luciana K. Calábria, Natalia F.Martins, Roberto Togawa, Foued S. Espindola, John R. Yates, Ricardo B. Cunha, and Marcelo V. de Sousa. Worker honeybee brain proteome. Journal of Proteome Research, 11(3):1485–1493, March 2012. ISSN 1535-3907. doi: 10.1021/ pr2007818.

[225] Selim Terhzaz, Pablo Cabrero, Robert A. Brinzer, Kenneth A. Halberg, Julian A. T. Dow, and Shireen-A. Davies. A novel role of Drosophila cytochrome P450-4e3 in permethrin insecticide tolerance. Insect Biochemistry and Molec- ular Biology, 67:38–46, December 2015. ISSN 0965-1748. doi: 10.1016/j.ibmb. 2015.06.002. URL http://www.sciencedirect.com/science/article/pii/ S0965174815300060.

[226] Cristina Manjon, Bartlomiej J. Troczka, Marion Zaworra, Katherine Beadle, Emma Randall, Gillian Hertlein, Kumar Saurabh Singh, Christoph T. Zimmer, Rafael A. Homem, Bettina Lueke, Rebecca Reid, Laura Kor, Maxie Kohler, Jürgen Benting, Martin S. Williamson, T. G. Emyr Davies, Linda M. Field, Chris Bass, and Ralf Nauen. Unravelling the Molecular Determinants of Bee Sensitivity to Neon- icotinoid Insecticides. Current Biology, 28(7):1137–1143.e5, April 2018. ISSN 0960-9822. doi: 10.1016/j.cub.2018.02.045. URL http://www.sciencedirect. com/science/article/pii/S0960982218302306.

[227] Bin Tang, Su Wang, Shi-Gui Wang, Hui-Juan Wang, Jia-Yong Zhang, and Shuai- Ying Cui. Invertebrate Trehalose-6-Phosphate Synthase Gene: Genetic Architec- ture, Biochemistry, Physiological Function, and Potential Applications. Frontiers

314 in Physiology, 9, January 2018. ISSN 1664-042X. doi: 10.3389/fphys.2018.00030. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5797772/.

[228] Hiroko Matsuda, Takayuki Yamada, Miki Yoshida, and Takashi Nishimura. Flies without Trehalose. Journal of Biological Chemistry, 290(2):1244–1255, January 2015. ISSN 0021-9258, 1083-351X. doi: 10.1074/jbc.M114.619411. URL http: //www.jbc.org/content/290/2/1244.

[229] Hyewon Chung, Ae-kyeong Kim, Sun-Ah Jung, Si Wouk Kim, Kweon Yu, and Joon H. Lee. The Drosophila homolog of methionine sulfoxide reductase A extends lifespan and increases nuclear localization of FOXO. FEBS Let- ters, 584(16):3609–3614, August 2010. ISSN 0014-5793. doi: 10.1016/j.febslet. 2010.07.033. URL http://www.sciencedirect.com/science/article/pii/ S0014579310005910.

[230] E.R. Stadtman, J. Moskovitz, B.S. Berlett, and R.L. Levine. Cyclic oxidation and reduction of protein methionine residues is an important antioxidant mechanism. Molecular and Cellular Biochemistry, 234-235:3–9, 2002. doi: 10.1023/A: 1015916831583.

[231] J. Moskovitz. Methionine sulfoxide reductases: Ubiquitous enzymes involved in antioxidant defense, protein regulation, and prevention of aging-associated diseases. Biochimica et Biophysica Acta - Proteins and Proteomics, 1703(2):213– 219, 2005. doi: 10.1016/j.bbapap.2004.09.003.

[232] R. C. Willis and C. A. Woolfolk. L-asparagine uptake in Escherichia coli. Journal of Bacteriology, 123(3):937–945, September 1975. ISSN 0021-9193, 1098-5530. URL http://jb.asm.org/content/123/3/937.

[233] Luis A. Clavell, Richard D. Gelber, Harvey J. Cohen, Suzanne Hitchcock-Bryan, J. Robert Cassady, Nancy J. Tarbell, Stephen R. Blattner, Ramana Tantravahi, Pearl Leavitt, and Stephen E. Sallan. Four-Agent Induction and Intensive As- paraginase Therapy for Treatment of Childhood Acute Lymphoblastic Leukemia.

315 New England Journal of Medicine, 315(11):657–663, September 1986. ISSN 0028- 4793. doi: 10.1056/NEJM198609113151101. URL https://doi.org/10.1056/ NEJM198609113151101.

[234] Verena Christen, Fabian Mittner, and Karl Fent. Molecular Effects of Neonicoti- noids in Honey Bees (Apis mellifera). Environmental Science & Technology, 50 (7):4071–4081, April 2016. ISSN 0013-936X. doi: 10.1021/acs.est.6b00678. URL https://doi.org/10.1021/acs.est.6b00678.

[235] Verena Christen, Melanie Schirrmann, Juerg E. Frey, and Karl Fent. Global Tran- scriptomic Effects of Environmentally Relevant Concentrations of the Neonicoti- noids Clothianidin, Imidacloprid, and Thiamethoxam in the Brain of Honey Bees ( Apis mellifera). Environmental Science & Technology, 52(13):7534–7544, July 2018. ISSN 1520-5851. doi: 10.1021/acs.est.8b01801.

[236] Zhiguo Li, Meng Li, Jingfang He, Xiaomeng Zhao, Veeranan Chaimanee, Wei- Fone Huang, Hongyi Nie, Yazhou Zhao, and Songkun Su. Differential physiological effects of neonicotinoid insecticides on honey bees: A comparison between Apis mellifera and Apis cerana. Pesticide Biochemistry and Physiology, 140:1–8, August 2017. ISSN 0048-3575. doi: 10.1016/j.pestbp.2017.06.010. URL http: //www.sciencedirect.com/science/article/pii/S0048357517300536.

[237] Nicola K. Simcock, Helen Gray, Soﬁa Bouchebti, and Geraldine A. Wright. Ap- petitive olfactory learning and memory in the honeybee depend on sugar re- ward identity. Journal of Insect Physiology, 106:71–77, April 2018. ISSN 0022- 1910. doi: 10.1016/j.jinsphys.2017.08.009. URL http://www.sciencedirect. com/science/article/pii/S0022191017301038.

[238] Sébastien C. Kessler, Erin Jo Tiedeken, Kerry L. Simcock, Sophie Derveau, Jes- sica Mitchell, Samantha Softley, Amy Radcliffe, Jane C. Stout, and Geraldine A. Wright. Bees prefer foods containing neonicotinoid pesticides. Nature, 521 (7550):74–76, May 2015. ISSN 1476-4687. doi: 10.1038/nature14414. URL https://www.nature.com/articles/nature14414. 316 [239] Roberto Bonasio, Qiye Li, Jinmin Lian, Navdeep S. Mutti, Lijun Jin, Hongmei Zhao, Pei Zhang, Ping Wen, Hui Xiang, Yun Ding, Zonghui Jin, Steven S. Shen, Zongji Wang, Wen Wang, Jun Wang, Shelley L. Berger, Jürgen Liebig, Guojie Zhang, and Danny Reinberg. Genome-wide and Caste-Speciﬁc DNA Methy- lomes of the Ants Camponotus ﬂoridanus and Harpegnathos saltator. Current Biology, 22(19):1755–1764, October 2012. ISSN 0960-9822. doi: 10.1016/j.cub. 2012.07.042. URL http://www.sciencedirect.com/science/article/pii/ S0960982212008676.

[240] Meromit Singer, Idit Kosti, Lior Pachter, and Yael Mandel-Gutfreund. A diverse epigenetic landscape at human exons with implication for expression. Nucleic Acids Research, 43(7):3498–3508, April 2015. ISSN 0305-1048. doi: 10.1093/nar/ gkv153. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4402514/.

[241] Brendan G. Hunt, Karl M. Glastad, Soojin V. Yi, and Michael A. D. Good- isman. The Function of Intragenic DNA Methylation: Insights from Insect Epigenomes. Integrative and Comparative Biology, 53(2):319–328, August 2013. ISSN 1540-7063. doi: 10.1093/icb/ict003. URL https://academic.oup.com/ icb/article/53/2/319/800483.

[242] Hume Stroud, Truman Do, Jiamu Du, Xuehua Zhong, Suhua Feng, Lianna John- son, Dinshaw J. Patel, and Steven E. Jacobsen. The roles of non-CG methylation in Arabidopsis. Nature structural & molecular biology, 21(1):64–72, January 2014. ISSN 1545-9993. doi: 10.1038/nsmb.2735. URL https://www.ncbi.nlm.nih. gov/pmc/articles/PMC4103798/.

[243] Olga V. Dyachenko, Tara V. Schevchuk, Leo Kretzner, Yaroslav I. Buryanov, and Steven S. Smith. Human non-CG methylation. Epigenetics, 5(7):569–572, Octo- ber 2010. ISSN 1559-2294. doi: 10.4161/epi.5.7.12702. URL https://doi.org/ 10.4161/epi.5.7.12702.

[244] Xu Wang, David Wheeler, Amanda Avery, Alfredo Rago, Jeong-Hyeon Choi, John K. Colbourne, Andrew G. Clark, and John H. Werren. Function and 317 Evolution of DNA Methylation in Nasonia vitripennis. PLOS Genetics, 9 (10):e1003872, October 2013. ISSN 1553-7404. doi: 10.1371/journal.pgen. 1003872. URL http://journals.plos.org/plosgenetics/article?id=10. 1371/journal.pgen.1003872.

[245] Paola Tognini, Debora Napoli, and Tommaso Pizzorusso. Dynamic DNA methylation in the brain: a new epigenetic mark for experience-dependent plasticity. Frontiers in Cellular Neuroscience, 9:331, 2015. ISSN 1662-5102. doi: 10.3389/fncel.2015.00331.

[246] K. M. Glastad, Brendan G. Hunt, S. V. Yi, and M. a. D. Goodisman. DNA methylation in insects: on the brink of the epigenomic era. Insect Molecular Biology, 20(5):553–565, October 2011. ISSN 1365-2583. doi: 10.1111/j.1365-2583.2011. 01092.x.

[247] Susan A. Weiner and Amy L. Toth. Epigenetics in Social Insects: A New Direction for Understanding the Evolution of Castes, 2012. URL https://www.hindawi. com/journals/gri/2012/609810/.

[248] Mat Welch and Ryan Lister. Epigenomics and the control of fate, form and function in social insects. Current Opinion in Insect Science, 1:31–38, July 2014. ISSN 2214-5745. doi: 10.1016/j.cois.2014.04.005. URL http://www.sciencedirect. com/science/article/pii/S2214574514000066.

[249] Mohan Manikkam, Carlos Guerrero-Bosagna, Rebecca Tracey, Md M. Haque, and Michael K. Skinner. Transgenerational Actions of Environmental Com- pounds on Reproductive Disease and Identiﬁcation of Epigenetic Biomarkers of Ancestral Exposures. PLOS ONE, 7(2):e31901, February 2012. ISSN 1932- 6203. doi: 10.1371/journal.pone.0031901. URL http://journals.plos.org/ plosone/article?id=10.1371/journal.pone.0031901.

[250] June L. Round and Sarkis K. Mazmanian. The gut microbiota shapes intestinal immune responses during health and disease. Nature Reviews Immunology, 9

318 (5):313–323, May 2009. ISSN 1474-1741. doi: 10.1038/nri2515. URL https: //www.nature.com/articles/nri2515.

[251] Philipp Engel and Nancy A. Moran. The gut microbiota of insects - diversity in structure and function. FEMS microbiology reviews, 37(5):699–735, September 2013. ISSN 1574-6976. doi: 10.1111/1574-6976.12025.

[252] Lora V. Hooper, Dan R. Littman, and Andrew J. Macpherson. Interactions between the microbiota and the immune system. Science (New York, N.Y.), 336 (6086):1268–1273, June 2012. ISSN 1095-9203. doi: 10.1126/science.1223490.

[253] Brendan A. Daisley, Mark Trinder, Tim W. McDowell, Stephanie L. Collins, Mark W. Sumarah, and Gregor Reid. Microbiota-mediated modulation of organophosphate insecticide toxicity by species-dependent lactobacilli interactions in a Drosophila melanogaster insect model. Appl. Environ. Micro- biol., pages AEM.02820–17, February 2018. ISSN 0099-2240, 1098-5336. doi: 10.1128/AEM.02820-17. URL https://aem.asm.org/content/early/2018/ 02/19/AEM.02820-17.

[254] Felix Sommer and Fredrik Bäckhed. The gut microbiota — masters of host development and physiology. Nature Reviews Microbiology, 11(4):227–238, April 2013. ISSN 1740-1534. doi: 10.1038/nrmicro2974. URL https://www.nature.com/ articles/nrmicro2974.

[255] J. Mrázek, L. Štrosová, K. Fliegerová, T. Kott, and J. Kopeˇcný. Diversity of insect intestinal microﬂora. Folia Microbiologica, 53(3):229–233, May 2008. ISSN 0015-5632, 1874-9356. doi: 10.1007/s12223-008-0032-z. URL https://link. springer.com/article/10.1007/s12223-008-0032-z.

[256] Brendan A. Daisley, Mark Trinder, Tim W. McDowell, Hylke Welle, Josh S. Dube, Sohrab N. Ali, Hon S. Leong, Mark W. Sumarah, and Gregor Reid. Neonicotinoid- induced pathogen susceptibility is mitigated by Lactobacillus plantarum immune stimulation in a Drosophila melanogaster model. Scientiﬁc Reports, 7

319 (1):2703, June 2017. ISSN 2045-2322. doi: 10.1038/s41598-017-02806-w. URL https://www.nature.com/articles/s41598-017-02806-w.

[257] Roberto Fusetto, Shane Denecke, Trent Perry, Richard A. J. O’Hair, and Philip Batterham. Partitioning the roles of CYP6g1 and gut microbes in the metabolism of the insecticide imidacloprid in Drosophila melanogaster. Scientiﬁc Reports, 7 (1):11339, September 2017. ISSN 2045-2322. doi: 10.1038/s41598-017-09800-2. URL https://www.nature.com/articles/s41598-017-09800-2.

[258] Karen M. Kapheim, Vikyath D. Rao, Carl J. Yeoman, Brenda A. Wilson, Bryan A. White, Nigel Goldenfeld, and Gene E. Robinson. Caste-Speciﬁc Differences in Hindgut Microbial Communities of Honey Bees (Apis mellifera). PLOS ONE, 10 (4):e0123911, April 2015. ISSN 1932-6203. doi: 10.1371/journal.pone.0123911. URL http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0123911.

[259] Nancy A. Moran, Allison K. Hansen, J. Elijah Powell, and Zakee L. Sabree. Dis- tinctive Gut Microbiota of Honey Bees Assessed Using Deep Sampling from In- dividual Worker Bees. PLOS ONE, 7(4):e36393, April 2012. ISSN 1932-6203. doi: 10.1371/journal.pone.0036393. URL http://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0036393.

[260] Nancy A. Moran. Genomics of the honey bee microbiome. Current opinion in insect science, 10:22–28, August 2015. ISSN 2214-5745. doi: 10.1016/j.cois.2015. 04.003. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4484875/.

[261] Waldan K. Kwong and Nancy A. Moran. Cultivation and characterization of the gut symbionts of honey bees and bumble bees: description of Snodgras- sella alvi gen. nov., sp. nov., a member of the family Neisseriaceae of the Be- taproteobacteria, and Gilliamella apicola gen. nov., sp. nov., a member of Or- baceae fam. nov., Orbales ord. nov., a sister taxon to the order ’Enterobacteri- ales’ of the Gammaproteobacteria. International Journal of Systematic and Evo- lutionary Microbiology, 63(Pt 6):2008–2018, June 2013. ISSN 1466-5034. doi: 320 10.1099/ijs.0.044875-0.

[262] Philipp Engel, Waldan K. Kwong, and Nancy A. Moran. Frischella perrara gen. nov., sp. nov., a gammaproteobacterium isolated from the gut of the honeybee, Apis mellifera. International Journal of Systematic and Evolutionary Microbi- ology, 63(Pt 10):3646–3651, October 2013. ISSN 1466-5034. doi: 10.1099/ijs.0. 049569-0.

[263] Waldan K. Kwong and Nancy A. Moran. Apibacter adventoris gen. nov., sp. nov., a member of the phylum Bacteroidetes isolated from honey bees. International Journal of Systematic and Evolutionary Microbiology, 66(3):1323–1329, March 2016. ISSN 1466-5034. doi: 10.1099/ijsem.0.000882.

[264] Hauke Koch and Paul Schmid-Hempel. Bacterial Communities in Central European Bumblebees: Low Diversity and High Speciﬁcity. Microbial Ecol- ogy, 62(1):121–133, July 2011. ISSN 0095-3628, 1432-184X. doi: 10.1007/ s00248-011-9854-3. URL https://link.springer.com/article/10.1007/ s00248-011-9854-3.

[265] Tobias C. Olofsson, Magnus Alsterfjord, Bo Nilson, Èile Butler, and Alejandra Vásquez. Lactobacillus apinorum sp. nov., Lactobacillus mellifer sp. nov., Lac- tobacillus mellis sp. nov., Lactobacillus melliventris sp. nov., Lactobacillus kimbladii sp. nov., Lactobacillus helsingborgensis sp. nov. and Lactobacillus kullabergensis sp. nov., isolated from the honey stomach of the honeybee Apis mellifera. International Journal of Systematic and Evolutionary Microbiology, 64(Pt 9):3109–3119, September 2014. ISSN 1466-5026. doi: 10.1099/ijs.0.059600-0. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4156108/.

[266] Vincent G. Martinson, Bryan N. Danforth, Robert L. Minckley, Olav Rueppell, Salim Tingek, and Nancy A. Moran. A simple and distinctive microbiota associated with honey bees and bumble bees. Molecular Ecology, 20(3):619–628, February 2011. ISSN 1365-294X. doi: 10.1111/j.1365-294X.2010.04959.x.

321 [267] J. Killer, J. Kopeˇcný, J. Mrázek, I. Koppová, J. Havlík, O. Benada, and T. Kott. Bi- ﬁdobacterium actinocoloniiforme sp. nov. and Biﬁdobacterium bohemicum sp. nov., from the bumblebee digestive tract. International Journal of Systematic and Evolutionary Microbiology, 61(Pt 6):1315–1321, June 2011. ISSN 1466-5034. doi: 10.1099/ijs.0.022525-0.

[268] J. Killer, J. Kopeˇcný, J. Mrázek, V. Rada, O. Benada, I. Koppová, J. Havlík, and J. Straka. Biﬁdobacterium bombi sp. nov., from the bumblebee digestive tract. International Journal of Systematic and Evolutionary Mi- crobiology, 59(8):2020–2024, 2009. doi: 10.1099/ijs.0.002915-0. URL http://ijs.microbiologyresearch.org/content/journal/ijsem/10. 1099/ijs.0.002915-0.

[269] Jessy Praet, Ivan Meeus, Margo Cnockaert, Kurt Houf, Guy Smagghe, and Peter Vandamme. Novel lactic acid bacteria isolated from the bumble bee gut: Conviv- ina intestini gen. nov., sp. nov., Lactobacillus bombicola sp. nov., and Weissella bombi sp. nov. Antonie Van Leeuwenhoek, 107(5):1337–1349, May 2015. ISSN 1572-9699. doi: 10.1007/s10482-015-0429-z.

[270] Tobias C. Olofsson and Alejandra Vásquez. Detection and identiﬁcation of a novel lactic acid bacterial ﬂora within the honey stomach of the honeybee Apis mellifera. Current Microbiology, 57(4):356–363, October 2008. ISSN 0343-8651. doi: 10.1007/s00284-008-9202-0.

[271] Kathrin I. Mohr and Christoph C. Tebbe. Diversity and phylotype consistency of bacteria in the guts of three bee species (Apoidea) at an oilseed rape ﬁeld. Environmental Microbiology, 8(2):258–272, February 2006. ISSN 1462-2912. doi: 10.1111/j.1462-2920.2005.00893.x.

[272] D. Arredondo, L. Castelli, M. P.Porrini, P.M. Garrido, M. J. Eguaras, P.Zunino, and K. Antúnez. Lactobacillus kunkeei strains decreased the infection by honey bee pathogens Paenibacillus larvae and Nosema ceranae. Beneﬁcial Microbes, 9(2): 279–290, February 2018. ISSN 1876-2891. doi: 10.3920/BM2017.0075. 322 [273] Blair K. Mockler, Waldan K. Kwong, Nancy A. Moran, and Hauke Koch. Micro- biome structure inﬂuences infection by the parasite Crithidia bombi in bumble bees. Applied and Environmental Microbiology, pages AEM.02335–17, Jan- uary 2018. ISSN 0099-2240, 1098-5336. doi: 10.1128/AEM.02335-17. URL http://aem.asm.org/content/early/2018/01/22/AEM.02335-17.

[274] W. G. Weisburg, S. M. Barns, D. A. Pelletier, and D. J. Lane. 16s ribosomal DNA ampliﬁcation for phylogenetic study. Journal of Bacteriology, 173(2):697–703, January 1991. ISSN 0021-9193.

[275] Filipe Pereira, João Carneiro, Rune Matthiesen, Barbara van Asch, Nádia Pinto, Leonor Gusmão, and António Amorim. Identiﬁcation of species by multiplex analysis of variable-length sequences. Nucleic Acids Research, 38(22):e203, De- cember 2010. ISSN 0305-1048. doi: 10.1093/nar/gkq865. URL https://www. ncbi.nlm.nih.gov/pmc/articles/PMC3001097/.

[276] Soumitesh Chakravorty, Danica Helb, Michele Burday, Nancy Connell, and David Alland. A detailed analysis of 16s ribosomal RNA gene segments for the diagnosis of pathogenic bacteria. Journal of microbiological methods, 69(2): 330–339, May 2007. ISSN 0167-7012. doi: 10.1016/j.mimet.2007.02.005. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2562909/.

[277] Bo Yang, Yong Wang, and Pei-Yuan Qian. Sensitivity and correlation of hypervariable regions in 16s rRNA genes in phylogenetic analysis. BMC Bioinformatics, 17 (1):135, March 2016. ISSN 1471-2105. doi: 10.1186/s12859-016-0992-y. URL https://doi.org/10.1186/s12859-016-0992-y.

[278] Kasie Raymann, Zack Shaffer, and Nancy A. Moran. Antibiotic exposure perturbs the gut microbiota and elevates mortality in honeybees. PLOS Biol- ogy, 15(3):e2001861, March 2017. ISSN 1545-7885. doi: 10.1371/journal.pbio. 2001861. URL http://journals.plos.org/plosbiology/article?id=10. 1371/journal.pbio.2001861.

323 [279] Erica Plummer, Jimmy Twin, Dieter M. Bulach, Suzanne M. Garl, and Sepehr N. Tabrizi. A Comparison of Three Bioinformatics Pipelines for the Analysis of Preterm Gut Microbiota using 16s rRNA Gene Sequencing Data. Journal of Proteomics & Bioinformatics, 8(12), December 2015. ISSN 0974-276X. doi: 10.4172/jpb.1000381. URL https://www.omicsonline.org/open-access/ a-comparison-of-three-bioinformatics-pipelines-for-the-analysis-ofpreterm-gut-microbiota-using-16s-rrna-gene-sequencing-data-jpb-1000381. php?aid=65142.

[280] Alexandre Almeida, Alex L. Mitchell, Aleksandra Tarkowska, and Robert D. Finn. Benchmarking taxonomic assignments based on 16s rRNA gene proﬁling of the microbiota from commonly sampled environments. GigaScience, 7(5), May 2018. doi: 10.1093/gigascience/giy054. URL https://academic.oup.com/ gigascience/article/7/5/giy054/4995265.

[281] Daniel McDonald, Morgan N. Price, Julia Goodrich, Eric P. Nawrocki, Todd Z. DeSantis, Alexander Probst, Gary L. Andersen, Rob Knight, and Philip Hugen- holtz. An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. The ISME Journal, 6(3): 610–618, March 2012. ISSN 1751-7370. doi: 10.1038/ismej.2011.139. URL https://www.nature.com/articles/ismej2011139.

[282] Scott Federhen. The NCBI Taxonomy database. Nucleic Acids Research, 40(D1): D136–D143, January 2012. ISSN 0305-1048. doi: 10.1093/nar/gkr1178. URL https://academic.oup.com/nar/article/40/D1/D136/2903327.

[283] Abigail Acland, Richa Agarwala, Tanya Barrett, Jeff Beck, Dennis A. Benson, Colleen Bollin, Evan Bolton, Stephen H. Bryant, Kathi Canese, Deanna M. Church, Karen Clark, Michael DiCuccio, Ilya Dondoshansky, Scott Federhen, Michael Feolo, Lewis Y. Geer, Viatcheslav Gorelenkov, Marilu Hoeppner, Mark Johnson, Christopher Kelly, Viatcheslav Khotomlianski, Avi Kimchi, Michael Kimelman, Paul Kitts, Sergey Krasnov, Anatoliy Kuznetsov, David Landsman, David J. Lipman, Zhiyong Lu, Thomas L. Madden, Tom Madej, Donna R. Maglott,

324 Aron Marchler-Bauer, Ilene Karsch-Mizrachi, Terence Murphy, James Ostell, Christopher O’Sullivan, Anna Panchenko, Lon Phan, Don Preussm Kim D. Pruitt, Wendy Rubinstein, Eric W. Sayers, Valerie Schneider, Gregory D. Schuler, Ed- win Sequeira, Stephen T. Sherry, Martin Shumway, Karl Sirotkin, Karanjit Siyan, Douglas Slotta, Alexandra Soboleva, Vladimir Soussov, Grigory Starchenko, Ta- tiana A. Tatusova, Bart W. Trawick, Denis Vakatov, Yanli Wang, Minghong Ward, W. John Wilbur, Eugene Yaschenko, and Kerry Zbicz. Database re- sources of the National Center for Biotechnology Information. Nucleic Acids Research, 42(Database issue):D7–D17, January 2014. ISSN 0305-1048. doi: 10.1093/nar/gkt1146. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC3965057/.

[284] James R. Cole, Qiong Wang, Jordan A. Fish, Benli Chai, Donna M. McGar- rell, Yanni Sun, C. Titus Brown, Andrea Porras-Alfaro, Cheryl R. Kuske, and James M. Tiedje. Ribosomal Database Project: data and tools for high throughput rRNA analysis. Nucleic Acids Research, 42(D1):D633–D642, January 2014. ISSN 0305-1048. doi: 10.1093/nar/gkt1244. URL https://academic.oup.com/nar/ article/42/D1/D633/1063201.

[285] Pelin Yilmaz, Laura Wegener Parfrey, Pablo Yarza, Jan Gerken, Elmar Pruesse, Christian Quast, Timmy Schweer, Jörg Peplies, Wolfgang Ludwig, and Frank Oliver Glöckner. The SILVA and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids Research, 42(D1):D643–D648, January 2014. ISSN 0305-1048. doi: 10.1093/nar/gkt1209. URL https://academic.oup.com/ nar/article/42/D1/D643/1061236.

[286] Cédric Alaux, Jean-Luc Brunet, Claudia Dussaubat, Fanny Mondet, Sylvie Tchamitchan, Marianne Cousin, Julien Brillard, Aurelie Baldy, Luc P.Belzunces, and Yves Le Conte. Interactions between Nosema microspores and a neonicotinoid weaken honeybees (Apis mellifera). Environmental Microbiology, 12 (3):774–782, 2010. ISSN 1462-2920. doi: 10.1111/j.1462-2920.2009.02123.x.

325 URL https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1462-2920. 2009.02123.x.

[287] Evan Bolyen, Jai Ram Rideout, Matthew R. Dillon, Nicholas A. Bokulich, Chris- tian Abnet, Gabriel A. Al-Ghalith, Harriet Alexander, Eric J. Alm, Manimozhiyan Arumugam, Francesco Asnicar, Yang Bai, Jordan E. Bisanz, Kyle Bittinger, Asker Brejnrod, Colin J. Brislawn, C. Titus Brown, Benjamin J. Callahan, Andrés Mauri- cio Caraballo-Rodríguez, John Chase, Emily Cope, Ricardo Da Silva, Pieter C. Dorrestein, Gavin M. Douglas, Daniel M. Durall, Claire Duvallet, Christian F.Ed- wardson, Madeleine Ernst, Mehrbod Estaki, Jennifer Fouquier, Julia M. Gauglitz, Deanna L. Gibson, Antonio Gonzalez, Kestrel Gorlick, Jiarong Guo, Benjamin Hillmann, Susan Holmes, Hannes Holste, Curtis Huttenhower, Gavin Huttley, Stefan Janssen, Alan K. Jarmusch, Lingjing Jiang, Benjamin Kaehler, Kyo Bin Kang, Christopher R. Keefe, Paul Keim, Scott T. Kelley, Dan Knights, Irina Koester, Tomasz Kosciolek, Jorden Kreps, Morgan GI Langille, Joslynn Lee, Ruth Ley, Yong-Xin Liu, Erikka Loftﬁeld, Catherine Lozupone, Massoud Maher, Clarisse Marotz, Bryan D. Martin, Daniel McDonald, Lauren J. McIver, Alexey V. Mel- nik, Jessica L. Metcalf, Sydney C. Morgan, Jamie Morton, Ahmad Turan Naimey, Jose A. Navas-Molina, Louis Felix Nothias, Stephanie B. Orchanian, Talima Pearson, Samuel L. Peoples, Daniel Petras, Mary Lai Preuss, Elmar Pruesse, Lasse Buur Rasmussen, Adam Rivers, I. I. Michael S Robeson, Patrick Rosen- thal, Nicola Segata, Michael Shaffer, Arron Shiffer, Rashmi Sinha, Se Jin Song, John R. Spear, Austin D. Swafford, Luke R. Thompson, Pedro J. Torres, Pauline Trinh, Anupriya Tripathi, Peter J. Turnbaugh, Sabah Ul-Hasan, Justin JJ van der Hooft, Fernando Vargas, Yoshiki Vázquez-Baeza, Emily Vogtmann, Max von Hip- pel, William Walters, Yunhu Wan, Mingxun Wang, Jonathan Warren, Kyle C. We- ber, Chase HD Williamson, Amy D. Willis, Zhenjiang Zech Xu, Jesse R. Zan- eveld, Yilong Zhang, Qiyun Zhu, Rob Knight, and J. Gregory Caporaso. QI- IME 2: Reproducible, interactive, scalable, and extensible microbiome data science. Technical Report e27295v2, PeerJ Preprints, December 2018. URL https:

326 //peerj.com/preprints/27295.

[288] Benjamin J Callahan, Paul J McMurdie, Michael J Rosen, Andrew W Han, Amy Jo A Johnson, and Susan P Holmes. DADA2: High resolution sample inference from Illumina amplicon data. Nature methods, 13(7):581–583, July 2016. ISSN 1548-7091. doi: 10.1038/nmeth.3869. URL https://www.ncbi.nlm.nih.gov/ pmc/articles/PMC4927377/.

[289] Catherine Lozupone and Rob Knight. UniFrac: a New Phylogenetic Method for Comparing Microbial Communities. Applied and Environmental Microbi- ology, 71(12):8228–8235, December 2005. ISSN 0099-2240. doi: 10.1128/AEM. 71.12.8228-8235.2005. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC1317376/.

[290] Kazutaka Katoh and Hiroyuki Toh. Recent developments in the MAFFT multiple sequence alignment program. Brieﬁngs in Bioinformatics, 9(4):286–298, July 2008. ISSN 1467-5463. doi: 10.1093/bib/bbn013. URL https://academic.oup. com/bib/article/9/4/286/266493.

[291] Animal and Plant Health Agency. Beebase - Beekeeping information resource for Beekeepers. URL http://www.nationalbeeunit.com/index.cfm? sectionid=26.

[292] Aline Fernanda Catae, Thaisa Cristina Roat, Marcel Pratavieira, Anally Ribeiro da Silva Menegasso, Mario Sergio Palma, and Osmar Malaspina. Exposure to a sublethal concentration of imidacloprid and the side effects on target and non- target organs of Apis mellifera (Hymenoptera, Apidae). Ecotoxicology (Lon- don, England), 27(2):109–121, March 2018. ISSN 1573-3017. doi: 10.1007/ s10646-017-1874-4.

[293] Jason A. Rothman, Mark J. Carroll, William G. Meikle, Kirk E. Anderson, and Quinn S. McFrederick. Longitudinal Effects of Supplemental Forage on the

327 Honey Bee (Apis mellifera) Microbiota and Inter- and Intra-Colony Variabil- ity. Microbial Ecology, February 2018. ISSN 1432-184X. doi: 10.1007/ s00248-018-1151-y. URL https://doi.org/10.1007/s00248-018-1151-y.

[294] Annelies Billiet, Ivan Meeus, Filip Van Nieuwerburgh, Dieter Deforce, Fe- lix Wäckers, and Guy Smagghe. Impact of sugar syrup and pollen diet on the bacterial diversity in the gut of indoor-reared bumblebees (Bombus terrestris). Apidologie, 47(4):548–560, July 2016. ISSN 1297-9678. doi: 10.1007/ s13592-015-0399-1. URL https://doi.org/10.1007/s13592-015-0399-1.

[295] Lucie Kešnerová, Ruben A. T. Mars, Kirsten M. Ellegaard, Michaël Troilo, Uwe Sauer, and Philipp Engel. Disentangling metabolic functions of bacteria in the honey bee gut. PLOS Biology, 15(12):e2003467, December 2017. ISSN 1545- 7885. doi: 10.1371/journal.pbio.2003467. URL https://journals.plos.org/ plosbiology/article?id=10.1371/journal.pbio.2003467.

[296] Ryan S. Schwarz, Nancy A. Moran, and Jay D. Evans. Early gut colonizers shape parasite susceptibility and microbiota composition in honey bee workers. Proceedings of the National Academy of Sciences, 113(33):9345–9350, Au- gust 2016. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.1606631113. URL http://www.pnas.org/content/113/33/9345.

[297] Érica Weinstein Teixeira, Lubiane Guimarães-Cestaro, Maria Luisa Teles Mar- ques Florêncio Alves, Dejair Message, Marta Fonseca Martins, Cynthia Fernan- des Pinto da Luz, and José Eduardo Serrão. Spores of Paenibacillus larvae, As- cosphaera apis, Nosema ceranae and Nosema apis in bee products supervised by the Brazilian Federal Inspection Service. Revista Brasileira de Entomologia, 62(3): 188–194, July 2018. ISSN 0085-5626. doi: 10.1016/j.rbe.2018.04.001. URL http: //www.sciencedirect.com/science/article/pii/S0085562617301760.

[298] Dave Goulson. Neonicotinoids impact bumblebee colony ﬁtness in the ﬁeld; a reanalysis of the UK’s Food & Environment Research Agency 2012 experiment.

328 PeerJ, 3:e854, March 2015. ISSN 2167-8359. doi: 10.7717/peerj.854. URL https: //peerj.com/articles/854.

[299] Michael C. Otterstatter and James D. Thomson. Does Pathogen Spillover from Commercially Reared Bumble Bees Threaten Wild Pollinators? PLOS ONE, 3(7):e2771, July 2008. ISSN 1932-6203. doi: 10.1371/journal. pone.0002771. URL https://journals.plos.org/plosone/article?id=10. 1371/journal.pone.0002771.

[300] Sydney A. Cameron, Jeffrey D. Lozier, James P. Strange, Jonathan B. Koch, Nils Cordes, Leellen F.Solter, and Terry L. Griswold. Patterns of widespread decline in North American bumble bees. Proceedings of the National Academy of Sciences, 108(2):662–667, January 2011. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas. 1014743108. URL http://www.pnas.org/content/108/2/662.

[301] M. A. Fürst, D. P. McMahon, J. L. Osborne, R. J. Paxton, and M. J. F. Brown. Dis- ease associations between honeybees and bumblebees as a threat to wild pollinators. Nature, 506(7488):364–366, February 2014. ISSN 1476-4687. doi: 10. 1038/nature12977. URL https://www.nature.com/articles/nature12977.

[302] Rachel E. Mallinger, Hannah R. Gaines-Day, and Claudio Gratton. Do managed bees have negative effects on wild bees?: A systematic review of the lit- erature. PLOS ONE, 12(12):e0189268, December 2017. ISSN 1932-6203. doi: 10.1371/journal.pone.0189268. URL https://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0189268.

[303] Javier Hernández López, Sophie Krainer, Antonia Engert, Wolfgang Schuehly, Ulrike Riessberger-Gallé, and Karl Crailsheim. Sublethal pesticide doses negatively affect survival and the cellular responses in American foulbrood-infected honeybee larvae. Scientiﬁc Reports, 7, February 2017. ISSN 2045-2322. doi: 10.1038/srep40853. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC5286422/.

329 [304] Aline Fauser-Misslin, Ben M. Sadd, Peter Neumann, and Christoph Sandrock. Inﬂuence of combined pesticide and parasite exposure on bumblebee colony traits in the laboratory. Journal of Applied Ecology, 51(2):450–459, April 2014. ISSN 1365-2664. doi: 10.1111/1365-2664.12188. URL https://besjournals. onlinelibrary.wiley.com/doi/abs/10.1111/1365-2664.12188.

[305] Louise Walderdorff, Philippe Laval-Gilly, Antoine Bonnefoy, and Jaïro Falla- Angel. Imidacloprid intensiﬁes its impact on honeybee and bumblebee cellular immune response when challenged with LPS (lippopolysacharide) of Es- cherichia coli. Journal of Insect Physiology, 108:17–24, July 2018. ISSN 0022- 1910. doi: 10.1016/j.jinsphys.2018.05.002. URL http://www.sciencedirect. com/science/article/pii/S0022191017304560.

[306] Eva Forsgren, Tobias C. Olofsson, Alejandra Váasquez, and Ingemar Fries. Novel lactic acid bacteria inhibiting Paenibacillus larvae in honey bee larvae. Apidolo- gie, 41(1):99–108, January 2010. ISSN 1297-9678. doi: 10.1051/apido/2009065. URL https://doi.org/10.1051/apido/2009065.

[307] Rachel S. Edgar, Edward W. Green, Yuwei Zhao, Gerben van Ooijen, Maria Olmedo, Ximing Qin, Yao Xu, Min Pan, Utham K. Valekunja, Kevin A. Feeney, Elizabeth S. Maywood, Michael H. Hastings, Nitin S. Baliga, Martha Merrow, Andrew J. Millar, Carl H. Johnson, Charalambos P. Kyriacou, John S. O’Neill, and Akhilesh B. Reddy. Peroxiredoxins are conserved markers of circadian rhythms. Nature, 485(7399):459–464, May 2012. ISSN 1476-4687. doi: 10.1038/ nature11088. URL https://www.nature.com/articles/nature11088.

[308] Santiago Mora-García, María José de Leone, and Marcelo Yanovsky. Time to grow: circadian regulation of growth and metabolism in photosynthetic organisms. Current Opinion in Plant Biology, 35:84–90, February 2017. ISSN 1369- 5266. doi: 10.1016/j.pbi.2016.11.009. URL http://www.sciencedirect.com/ science/article/pii/S1369526616301972.

330 [309] Guy Bloch, Noam Bar-Shai, Yotam Cytter, and Rachel Green. Time is honey: circadian clocks of bees and ﬂowers and how their interactions may inﬂuence ecological communities. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 372(1734), November 2017. ISSN 1471-2970. doi: 10.1098/rstb.2016.0256.

[310] Guy Bloch. The Social Clock of the Honeybee. Journal of Biological Rhythms, 25(5):307–317, October 2010. ISSN 0748-7304. doi: 10.1177/0748730410380149. URL https://doi.org/10.1177/0748730410380149.

[311] T. Hovestadt, T. Degen, and O. Mitesser. Suitable triggers for timing the transition from worker to sexual production in annual eusocial insects. Insectes Sociaux, 65 (4):609–617, November 2018. ISSN 1420-9098. doi: 10.1007/s00040-018-0652-5. URL https://doi.org/10.1007/s00040-018-0652-5.

[312] G. Bloch, C. D. Rubinstein, and G. E. Robinson. period expression in the honey bee brain is developmentally regulated and not affected by light, ﬂight experience, or colony type. Insect Biochemistry and Molecular Biology, 34(9):879–891, September 2004. ISSN 0965-1748. doi: 10.1016/j.ibmb.2004.05.004.

[313] Thomas A. Sebeok. Review of Communication among social bees, ; Porpoises and Sonar, ; Man and Dolphin. Language, 39(3):448–466, 1963. ISSN 0097-8507. doi: 10.2307/411126. URL https://www.jstor.org/stable/411126.

[314] Thomas D. Seeley. The tremble dance of the honey bee: message and meanings. Behavioral Ecology and Sociobiology, 31(6):375–383, December 1992. ISSN 1432- 0762. doi: 10.1007/BF00170604. URL https://doi.org/10.1007/BF00170604.

[315] Mickaël Henry, Maxime Béguin, Fabrice Requier, Orianne Rollin, Jean-François Odoux, Pierrick Aupinel, Jean Aptel, Sylvie Tchamitchian, and Axel Decour- tye. A Common Pesticide Decreases Foraging Success and Survival in Honey Bees. Science, 336(6079):348–350, April 2012. ISSN 0036-8075, 1095-9203. doi:

331 10.1126/science.1215039. URL http://science.sciencemag.org/content/ 336/6079/348.

[316] Ezio Rosato and Charalambos P. Kyriacou. Analysis of locomotor activity rhythms in Drosophila. Nature Protocols, 1(2):559–568, August 2006. ISSN 1750- 2799. doi: 10.1038/nprot.2006.79. URL https://www.nature.com/articles/ nprot.2006.79.

[317] Eric C. R. Reeve and Isobel Black. Encyclopedia of Genetics. Taylor & Francis, 2001. ISBN 9781884964343. Google-Books-ID: JjLWYKqehRsC.

[318] Azusa Kamikouchi, Hidehiko K. Inagaki, Thomas Effertz, Oliver Hendrich, André Fiala, Martin C. Göpfert, and Kei Ito. The neural basis of Drosophila gravity- sensing and hearing. Nature, 458(7235):165–171, March 2009. ISSN 1476- 4687. doi: 10.1038/nature07810. URL https://www.nature.com/articles/ nature07810.

[319] Seymour Benzer. BEHAVIORAL MUTANTS OF Drosophila ISOLATED BY COUN- TERCURRENT DISTRIBUTION. Proceedings of the National Academy of Sci- ences, 58(3):1112–1119, September 1967. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.58.3.1112. URL http://www.pnas.org/content/58/3/1112.

[320] Thomas J. Colgan, Isabel K. Fletcher, Andres N. Arce, Richard J. Gill, Ana Ramos Rodrigues, Eckart Stolle, Lars Chittka, and Yannick Wurm. Caste- and pesticide- speciﬁc effects of neonicotinoid pesticide exposure on gene expression in bumblebees. Molecular Ecology, 0(0). ISSN 1365-294X. doi: 10.1111/mec.15047. URL https://onlinelibrary.wiley.com/doi/abs/10.1111/mec.15047.

[321] Catherine E. Schretter, Jost Vielmetter, Imre Bartos, Zsuzsa Marka, Szabolcs Marka, Sulabha Argade, and Sarkis K. Mazmanian. A gut microbial factor modulates locomotor behaviour in Drosophila. Nature, 563(7731):402, November 2018. ISSN 1476-4687. doi: 10.1038/s41586-018-0634-9. URL https://www. nature.com/articles/s41586-018-0634-9.

332 [322] Abdessalam Kacimi El Hassani, Matthieu Dacher, Vincent Gary, Michel Lambin, Monique Gauthier, and Catherine Armengaud. Effects of sublethal doses of acetamiprid and thiamethoxam on the behavior of the honeybee (Apis mellifera). Archives of Environmental Contamination and Toxicology, 54(4):653–661, May 2008. ISSN 1432-0703. doi: 10.1007/s00244-007-9071-8.

[323] Sally M. Williamson, Sarah J. Willis, and Geraldine A. Wright. Exposure to neonicotinoids inﬂuences the motor function of adult worker honeybees. Eco- toxicology, 23(8):1409–1418, October 2014. ISSN 1573-3017. doi: 10.1007/ s10646-014-1283-x. URL https://doi.org/10.1007/s10646-014-1283-x.

[324] Julien Dupuis, Thierry Louis, Monique Gauthier, and Valérie Raymond. In- sights from honeybee (Apis mellifera) and ﬂy (Drosophila melanogaster) nicotinic acetylcholine receptors: From genes to behavioral functions. Neuro- science & Biobehavioral Reviews, 36(6):1553–1564, July 2012. ISSN 0149-7634. doi: 10.1016/j.neubiorev.2012.04.003. URL http://www.sciencedirect.com/ science/article/pii/S0149763412000620.

[325] Elad B. Rubin, Yair Shemesh, Mira Cohen, Sharona Elgavish, Hugh M. Robert- son, and Guy Bloch. Molecular and phylogenetic analyses reveal mammalian- like clockwork in the honey bee (Apis mellifera) and shed new light on the molecular evolution of the circadian clock. Genome Research, 16(11):000–000, November 2006. ISSN 1088-9051, 1549-5469. doi: 10.1101/gr.5094806. URL http://genome.cshlp.org/content/early/2006/10/25/gr.5094806.

[326] Louisa A. Hooven, Katherine A. Sherman, Shawn Butcher, and Jadwiga M. Giebultowicz. Does the Clock Make the Poison? Circadian Variation in Re- sponse to Pesticides. PLOS ONE, 4(7):e6469, July 2009. ISSN 1932-6203. doi: 10.1371/journal.pone.0006469. URL https://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0006469.

333 [327] Francis Levi and Ueli Schibler. Circadian rhythms: mechanisms and therapeutic implications. Annual Review of Pharmacology and Toxicology, 47:593–628, 2007. ISSN 0362-1642. doi: 10.1146/annurev.pharmtox.47.120505.105208.

[328] Curt Stern and Elizabeth White Schaeffer. On Primary Attributes of Alleles in Drosophila Melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 29(11):351–361, December 1943. ISSN 0027-8424. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1078632/.

[329] Benjamin Schmid, Charlotte Helfrich-Förster, and Taishi Yoshii. A New Im- ageJ Plug-in “ActogramJ” for Chronobiological Analyses. Journal of Biologi- cal Rhythms, 26(5):464–467, October 2011. ISSN 0748-7304. doi: 10.1177/ 0748730411414264. URL https://doi.org/10.1177/0748730411414264.

[330] K. V. Allebrandt, N. Amin, B. Müller-Myhsok, T. Esko, M. Teder-Laving, R. V. D. M. Azevedo, C. Hayward, J. van Mill, N. Vogelzangs, E. W. Green, S. A. Melville, P. Lichtner, H.-E. Wichmann, B. A. Oostra, A. C. J. W. Janssens, H. Campbell, J. F. Wilson, A. A. Hicks, P.P.Pramstaller, Z. Dogas, I. Rudan, M. Merrow, B. Pen- ninx, C. P. Kyriacou, A. Metspalu, C. M. van Duijn, T. Meitinger, and T. Roen-

neberg. A Katp channel gene effect on sleep duration: from genome-wide association studies to function in Drosophila. Molecular Psychiatry, 18(1):122– 132, January 2013. ISSN 1476-5578. doi: 10.1038/mp.2011.142. URL https: //www.nature.com/articles/mp2011142.

[331] Nina Pekkala, K. Emily Knott, Janne S. Kotiaho, Kari Nissinen, and Mikael Pu- urtinen. The effect of inbreeding rate on ﬁtness, inbreeding depression and heterosis over a range of inbreeding coefﬁcients. Evolutionary Applications, 7 (9):1107–1119, November 2014. ISSN 1752-4571. doi: 10.1111/eva.12145. URL https://onlinelibrary.wiley.com/doi/abs/10.1111/eva.12145.

[332] Zachariah P. G. Olufs, Carin A. Loewen, Barry Ganetzky, David A. Wassarman, and Misha Perouansky. Genetic variability affects absolute and relative poten- cies and kinetics of the anesthetics isoflurane and sevoflurane in Drosophila 334 melanogaster. Scientific Reports, 8(1):2348, February 2018. ISSN 2045-2322. doi: 10.1038/s41598-018-20720-7. URL https://www.nature.com/articles/ s41598-018-20720-7.

[333] Alexandra G. Imasheva, Dmitri V. Bosenko, and Oleg A. Bubli. Variation in morphological traits of Drosophila melanogaster (fruit ﬂy) under nutritional stress. Heredity, 82(2):187–192, February 1999. ISSN 1365-2540. doi: 10.1046/j. 1365-2540.1999.00466.x. URL https://onlinelibrary.wiley.com/doi/abs/ 10.1046/j.1365-2540.1999.00466.x.

[334] Brian P. Lazzaro, Timothy B. Sackton, and Andrew G. Clark. Genetic Variation in Drosophila melanogaster Resistance to Infection: A Comparison Across Bac- teria. Genetics, 174(3):1539–1554, November 2006. ISSN 0016-6731, 1943-2631. doi: 10.1534/genetics.105.054593. URL http://www.genetics.org/content/ 174/3/1539.

[335] David Duneau, Haina Sun, Jonathan Revah, Keri San Miguel, Henry D. Kunerth, Ian V. Caldas, Philipp W. Messer, Jeffrey G. Scott, and Nicolas Buchon. Signatures of Insecticide Selection in the Genome of Drosophila melanogaster. G3: Genes, Genomes, Genetics, 8(11):3469–3480, November 2018. ISSN 2160-1836. doi: 10. 1534/g3.118.200537. URL http://www.g3journal.org/content/8/11/3469.

[336] Thomas WR Harrop, Shane Denecke, Ying Ting Yang, Janice Chan, Phillip J. Daborn, Trent Perry, and Philip Batterham. Evidence for activation of niten- pyram by a mitochondrial cytochrome P450 in Drosophila melanogaster. Pest Management Science, 74(7):1616–1622, July 2018. ISSN 1526-4998. doi: 10. 1002/ps.4852. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/ ps.4852.

[337] Phillip J. Daborn, Christopher Lumb, Adrian Boey, Wayn Wong, Richard H. ffrench Constant, and Philip Batterham. Evaluating the insecticide resistance

335 potential of eight Drosophila melanogaster cytochrome P450 genes by trans- genic over-expression. Insect Biochemistry and Molecular Biology, 37(5):512– 519, May 2007. ISSN 0965-1748. doi: 10.1016/j.ibmb.2007.02.008. URL http: //www.sciencedirect.com/science/article/pii/S0965174807000501.

[338] Laura Michelle Beaver, Louisa Ada Hooven, Shawn Michael Butcher, Natraj Kr- ishnan, Katherine Alice Sherman, Eileen Shin-Yeu Chow, and Jadwiga Maria Giebultowicz. Circadian Clock Regulates Response to Pesticides in Drosophila via Conserved Pdp1 Pathway. Toxicological Sciences, 115(2):513–520, June 2010. ISSN 1096-6080. doi: 10.1093/toxsci/kfq083. URL https://academic.oup. com/toxsci/article/115/2/513/1736133.

[339] Michal Krzyzanowski. WHO Air Quality Guidelines for Europe. Journal of Tox- icology and Environmental Health, Part A, 71(1):47–50, November 2008. ISSN 1528-7394. doi: 10.1080/15287390701557834. URL https://doi.org/10. 1080/15287390701557834.

[340] Aaron J. Cohen, H. Ross Anderson, Bart Ostro, Kiran Dev Pandey, Michal Krzyzanowski, Nino Künzli, Kersten Gutschmidt, Arden Pope, Isabelle Romieu, Jonathan M. Samet, and Kirk Smith. The Global Burden of Disease Due to Out- door Air Pollution. Journal of Toxicology and Environmental Health, Part A, 68 (13-14):1301–1307, July 2005. ISSN 1528-7394. doi: 10.1080/15287390590936166. URL https://doi.org/10.1080/15287390590936166.

[341] Alistair Thorpe and Roy M. Harrison. Sources and properties of non-exhaust particulate matter from road trafﬁc: A review. Science of The Total Environ- ment, 400(1):270–282, August 2008. ISSN 0048-9697. doi: 10.1016/j.scitotenv. 2008.06.007. URL http://www.sciencedirect.com/science/article/pii/ S004896970800658X.

[342] Ki-Hyun Kim, Ehsanul Kabir, and Shamin Kabir. A review on the human health impact of airborne particulate matter. Environment International, 74:136–143,

336 January 2015. ISSN 0160-4120. doi: 10.1016/j.envint.2014.10.005. URL http: //www.sciencedirect.com/science/article/pii/S0160412014002992.

[343] Frank J. Kelly and Julia C. Fussell. Size, source and chemical composition as determinants of toxicity attributable to ambient particulate matter. Atmospheric Environment, 60:504–526, December 2012. ISSN 1352-2310. doi: 10.1016/j.atmosenv.2012.06.039. URL http://www.sciencedirect.com/ science/article/pii/S1352231012006024.

[344] F Laden, L M Neas, D W Dockery, and J Schwartz. Association of ﬁne particulate matter from different sources with daily mortality in six U.S. cities. Environmen- tal Health Perspectives, 108(10):941–947, October 2000. ISSN 0091-6765. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1240126/.

[345] Mathew R. Heal, Prashant Kumar, and Roy M. Harrison. Particles, air quality, policy and health. Chemical Society Reviews, 41(19):6606–6630, 2012. doi: 10.1039/ C2CS35076A. URL https://pubs.rsc.org/en/content/articlelanding/ 2012/cs/c2cs35076a.

[346] Roberto Casati, Volker Scheer, Rainer Vogt, and Thorsten Benter. Measure- ment of nucleation and soot mode particle emission from a diesel passen- ger car in real world and laboratory in situ dilution. Atmospheric Environ- ment, 41(10):2125–2135, March 2007. ISSN 1352-2310. doi: 10.1016/j.atmosenv. 2006.10.078. URL http://www.sciencedirect.com/science/article/pii/ S1352231006010983.

[347] D Butterﬁeld, S Beccaceci, P Quincey, B Sweeney, A Lilley, C Bradshaw, G Fuller, D Green, and A Font. 2015 Annual report for the UK Black Carbon Network. : Publications : National Physical Laboratory, 2015. URL http://www.npl.co. uk/content/ConPublication/7280.

[348] Neeta Kulkarni, Nevil Pierse, Lesley Rushton, and Jonathan Grigg. Carbon in Airway Macrophages and Lung Function in Children. New England Journal of

337 Medicine, 355(1):21–30, July 2006. ISSN 0028-4793. doi: 10.1056/NEJMoa052972. URL https://doi.org/10.1056/NEJMoa052972.

[349] Weiwei Lin, Wei Huang, Tong Zhu, Min Hu, Bert Brunekreef, Yuanhang Zhang, Xingang Liu, Hong Cheng, Ulrike Gehring, Chengcai Li, and Xiaoyan Tang. Acute Respiratory Inﬂammation in Children and Black Carbon in Ambient Air before and during the 2008 Beijing Olympics. Environmental Health Perspectives, 119 (10):1507–1512, October 2011. ISSN 0091-6765. doi: 10.1289/ehp.1103461. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3230448/.

[350] Philipp Engel, Vincent G. Martinson, and Nancy A. Moran. Functional diversity within the simple gut microbiota of the honey bee. Proceedings of the National Academy of Sciences, page 201202970, June 2012. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.1202970109. URL http://www.pnas.org/content/early/ 2012/06/13/1202970109.

[351] Germán Bonilla-Rosso and Philipp Engel. Functional roles and metabolic niches in the honey bee gut microbiota. Current Opinion in Microbiology, 43:69–76, June 2018. ISSN 1369-5274. doi: 10.1016/j.mib.2017.12.009. URL http://www. sciencedirect.com/science/article/pii/S1369527417300693.

[352] Erick V. S. Motta, Kasie Raymann, and Nancy A. Moran. Glyphosate perturbs the gut microbiota of honey bees. Proceedings of the National Academy of Sciences, page 201803880, September 2018. ISSN 0027-8424, 1091-6490. doi: 10.1073/ pnas.1803880115. URL http://www.pnas.org/content/early/2018/09/18/ 1803880115.

[353] Kathrin Näpﬂin and Paul Schmid-Hempel. High Gut Microbiota Diversity Pro- vides Lower Resistance against Infection by an Intestinal Parasite in Bumble- bees. The American Naturalist, 192(2):131–141, June 2018. ISSN 0003-0147. doi: 10.1086/698013. URL https://www.journals.uchicago.edu/doi/abs/ 10.1086/698013.

338 [354] Stephen A. Bustin, Vladimir Benes, Jeremy A. Garson, Jan Hellemans, Jim Huggett, Mikael Kubista, Reinhold Mueller, Tania Nolan, Michael W. Pfafﬂ, Gre- gory L. Shipley, Jo Vandesompele, and Carl T. Wittwer. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experi- ments. Clinical Chemistry, 55(4):611–622, April 2009. ISSN 0009-9147, 1530- 8561. doi: 10.1373/clinchem.2008.112797. URL http://clinchem.aaccjnls. org/content/55/4/611.

[355] Marie Louise Hermann-Bank, Kerstin Skovgaard, Anders Stockmarr, Niels Larsen, and Lars Mølbak. The Gut Microbiotassay: a high-throughput qPCR approach combinable with next generation sequencing to study gut microbial diversity. BMC Genomics, 14(1):788, November 2013. ISSN 1471- 2164. doi: 10.1186/1471-2164-14-788. URL https://doi.org/10.1186/ 1471-2164-14-788.

[356] Noah Fierer, Jason A. Jackson, Rytas Vilgalys, and Robert B. Jackson. Assess- ment of Soil Microbial Community Structure by Use of Taxon-Speciﬁc Quanti- tative PCR Assays. Appl. Environ. Microbiol., 71(7):4117–4120, July 2005. ISSN 0099-2240, 1098-5336. doi: 10.1128/AEM.71.7.4117-4120.2005. URL https: //aem.asm.org/content/71/7/4117.

[357] Ivan Meeus, Laurian Parmentier, Annelies Billiet, Kevin Maebe, Filip Van Nieuwerburgh, Dieter Deforce, Felix Wäckers, Peter Vandamme, and Guy Smag- ghe. 16s rRNA Amplicon Sequencing Demonstrates that Indoor-Reared Bumble- bees (Bombus terrestris) Harbor a Core Subset of Bacteria Normally Associated with the Wild Host. PLOS ONE, 10(4):e0125152, April 2015. ISSN 1932-6203. doi: 10.1371/journal.pone.0125152. URL https://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0125152.

[358] Michael W. Pfafﬂ. A new mathematical model for relative quantiﬁcation in real- time RT–PCR. Nucleic Acids Research, 29(9):e45, May 2001. ISSN 0305-1048. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC55695/.

339 [359] Vanessa Corby-Harris, Patrick Maes, and Kirk E. Anderson. The Bacterial Communities Associated with Honey Bee (Apis mellifera) Foragers. PLOS ONE, 9(4):e95056, April 2014. ISSN 1932-6203. doi: 10.1371/journal. pone.0095056. URL https://journals.plos.org/plosone/article?id=10. 1371/journal.pone.0095056.

[360] J. Lelieveld, J. S. Evans, M. Fnais, D. Giannadaki, and A. Pozzer. The contribution of outdoor air pollution sources to premature mortality on a global scale. Nature, 525(7569):367–371, September 2015. ISSN 1476-4687. doi: 10.1038/nature15371. URL https://www.nature.com/articles/nature15371.

[361] Shane G. DuBay and Carl C. Fuldner. Bird specimens track 135 years of atmospheric black carbon and environmental policy. Proceedings of the National Academy of Sciences, 114(43):11321–11326, October 2017. ISSN 0027-8424, 1091- 6490. doi: 10.1073/pnas.1710239114. URL http://www.pnas.org/content/ 114/43/11321.

[362] Ece A. Mutlu, I¸sınY. Comba, Takugo Cho, Phillip A. Engen, Cemal Yazıcı, Saul Soberanes, Robert B. Hamanaka, Recep Ni˘gdelio˘glu, Angelo Y. Meliton, An- drew J. Ghio, G. R. Scott Budinger, and Gökhan M. Mutlu. Inhalational exposure to particulate matter air pollution alters the composition of the gut microbiome. Environmental Pollution, 240:817–830, September 2018. ISSN 0269- 7491. doi: 10.1016/j.envpol.2018.04.130. URL http://www.sciencedirect. com/science/article/pii/S0269749117351552.

[363] Leigh A. Beamish, Alvaro R. Osornio-Vargas, and Eytan Wine. Air pollution: An environmental factor contributing to intestinal disease. Journal of Crohn’s and Colitis, 5(4):279–286, August 2011. ISSN 1873-9946. doi: 10.1016/j. crohns.2011.02.017. URL https://academic.oup.com/ecco-jcc/article/ 5/4/279/653514.

[364] Xuechao Guo, Su Liu, Zhu Wang, Xu-xiang Zhang, Mei Li, and Bing Wu. Metage- nomic proﬁles and antibiotic resistance genes in gut microbiota of mice exposed 340 to arsenic and iron. Chemosphere, 112:1–8, October 2014. ISSN 1879-1298. doi: 10.1016/j.chemosphere.2014.03.068.

[365] Xu Han, Brennen Geller, Kristy Moniz, Pranab Das, Adam K. Chippindale, and Virginia K. Walker. Monitoring the developmental impact of copper and sil- ver nanoparticle exposure in Drosophila and their microbiomes. The Science of the Total Environment, 487:822–829, July 2014. ISSN 1879-1026. doi: 10.1016/ j.scitotenv.2013.12.129.

[366] M. Carina Audisio, María J. Torres, Daniela C. Sabaté, Carolina Ibarguren, and María C. Apella. Properties of different lactic acid bacteria isolated from Apis mellifera L. bee-gut. Microbiological Research, 166(1):1–13, January 2011. ISSN 0944-5013. doi: 10.1016/j.micres.2010.01.003. URL http://www. sciencedirect.com/science/article/pii/S0944501310000042.

[367] Annelies Billiet, Ivan Meeus, Filip Van Nieuwerburgh, Dieter Deforce, Felix Wäckers, and Guy Smagghe. Colony contact contributes to the diversity of gut bacteria in bumblebees (Bombus terrestris). Insect Science, 24(2):270–277, April 2017. ISSN 1744-7917. doi: 10.1111/1744-7917.12284. URL https: //onlinelibrary.wiley.com/doi/abs/10.1111/1744-7917.12284.

[368] THE EUROPEAN COMMISSION. EUR-Lex - 32013r0485 - EN - EUR-Lex, 2013. URL https://eur-lex.europa.eu/eli/reg_impl/2013/485/oj.

[369] Wenfu Mao, Mary A. Schuler, and May R. Berenbaum. CYP9q-mediated detox- iﬁcation of acaricides in the honey bee (Apis mellifera). Proceedings of the National Academy of Sciences of the United States of America, 108(31):12657– 12662, August 2011. ISSN 0027-8424. doi: 10.1073/pnas.1109535108. URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3150950/.

341 [370] Johannes Fischer, Teresa Müller, Anne-Kathrin Spatz, Uwe Greggers, Bernd Grünewald, and Randolf Menzel. Neonicotinoids interfere with speciﬁc compo- nents of navigation in honeybees. PloS One, 9(3):e91364, 2014. ISSN 1932-6203. doi: 10.1371/journal.pone.0091364.

[371] Ken Tan, Weiwen Chen, Shihao Dong, Xiwen Liu, Yuchong Wang, and James C. Nieh. Imidacloprid Alters Foraging and Decreases Bee Avoidance of Predators. PLOS ONE, 9(7):e102725, July 2014. ISSN 1932-6203. doi: 10.1371/journal. pone.0102725. URL https://journals.plos.org/plosone/article?id=10. 1371/journal.pone.0102725.

[372] Léa Tison, Marie-Luise Hahn, Sophie Holtz, Alexander Rößner, Uwe Greggers, Gabriela Bischoff, and Randolf Menzel. Honey Bees’ Behavior Is Impaired by Chronic Exposure to the Neonicotinoid Thiacloprid in the Field. Environmental Science & Technology, 50(13):7218–7227, July 2016. ISSN 0013-936X. doi: 10. 1021/acs.est.6b02658. URL https://doi.org/10.1021/acs.est.6b02658.

[373] Tobias Pamminger, Christina Botías, Dave Goulson, and William O. H. Hughes. A mechanistic framework to explain the immunosuppressive effects of neurotoxic pesticides on bees. Functional Ecology, 32(8):1921–1930, August 2018. ISSN 1365-2435. doi: 10.1111/1365-2435.13119. URL https://besjournals. onlinelibrary.wiley.com/doi/abs/10.1111/1365-2435.13119.

[374] Gennaro Di Prisco, Valeria Cavaliere, Desiderato Annoscia, Paola Varricchio, Emilio Caprio, Francesco Nazzi, Giuseppe Gargiulo, and Francesco Pennac- chio. Neonicotinoid clothianidin adversely affects insect immunity and pro- motes replication of a viral pathogen in honey bees. Proceedings of the Na- tional Academy of Sciences, 110(46):18466–18471, November 2013. ISSN 0027- 8424, 1091-6490. doi: 10.1073/pnas.1314923110. URL http://www.pnas.org/ content/110/46/18466.

[375] Arthur David, Cristina Botías, Alaa Abdul-Sada, Elizabeth Nicholls, Ellen L. Rotheray, Elizabeth M. Hill, and Dave Goulson. Widespread contamination of 342 wildﬂower and bee-collected pollen with complex mixtures of neonicotinoids and fungicides commonly applied to crops. Environment International, 88:169– 178, March 2016. ISSN 0160-4120. doi: 10.1016/j.envint.2015.12.011. URL http: //www.sciencedirect.com/science/article/pii/S0160412015301161.

[376] Franziska Böhme, Gabriela Bischoff, Claus PW Zebitz, Peter Rosenkranz, and Klaus Wallner. Chronic exposure of honeybees, Apis mellifera (Hymenoptera: Apidae), to a pesticide mixture in realistic ﬁeld exposure rates. Apidologie, 48 (3):353–363, 2017. ISSN 1297-9678. doi: 10.1007/s13592-016-0479-x. URL https://doi.org/10.1007/s13592-016-0479-x.

[377] Fabio Sgolastra, Xavier Arnan, Riccardo Cabbri, Gloria Isani, Piotr Medrzycki, Dariusz Teper, and Jordi Bosch. Combined exposure to sublethal concentrations of an insecticide and a fungicide affect feeding, ovary development and longevity in a solitary bee. Proc. R. Soc. B, 285(1885):20180887, August 2018. ISSN 0962-8452, 1471-2954. doi: 10.1098/rspb.2018.0887. URL http://rspb. royalsocietypublishing.org/content/285/1885/20180887.

[378] Fabio Sgolastra, Piotr Medrzycki, Laura Bortolotti, Maria Teresa Renzi, Simone Tosi, Gherardo Bogo, Dariusz Teper, Claudio Porrini, Roberto Molowny-Horas, and Jordi Bosch. Synergistic mortality between a neonicotinoid insecticide and an ergosterol-biosynthesis-inhibiting fungicide in three bee species. Pest Man- agement Science, 73(6):1236–1243, June 2017. ISSN 1526-4998. doi: 10.1002/ps. 4449.

[379] N. Simon-Delso, V. Amaral-Rogers, L. P.Belzunces, J. M. Bonmatin, M. Chagnon, C. Downs, L. Furlan, D. W. Gibbons, C. Giorio, V. Girolami, D. Goulson, D. P. Kreutzweiser, C. H. Krupke, M. Liess, E. Long, M. McField, P. Mineau, E. A. D. Mitchell, C. A. Morrissey, D. A. Noome, L. Pisa, J. Settele, J. D. Stark, A. Tapparo, H. Van Dyck, J. Van Praagh, J. P.Van der Sluijs, P.R. Whitehorn, and M. Wiemers. Systemic insecticides (neonicotinoids and ﬁpronil): trends, uses, mode of action and metabolites. Environmental Science and Pollution Research, 22(1):

343 5–34, January 2015. ISSN 1614-7499. doi: 10.1007/s11356-014-3470-y. URL https://doi.org/10.1007/s11356-014-3470-y.

[380] Health Canada and Health Canada. Registration Decision RD2016-12, Sul- foxaﬂor, April 2016. URL https://www.canada.ca/en/health-canada/ services/consumer-product-safety/reports-publications/ pesticides-pest-management/decisions-updates/ registration-decision/2016/sulfoxaflor-rd2016-12.html.

[381] Australian Pesticides and Veterinary Medicines Authority. Gazette No. 13, 2 July 2013, June 2014. URL https://apvma.gov.au/node/10401.

[382] European Food Safety Authority. Peer review of the pesticide risk assessment of the active substance sulfoxaﬂor, May 2014. URL https://www.efsa.europa. eu/en/efsajournal/pub/3692.

[383] Thomas C. Sparks, Gerald B. Watson, Michael R. Loso, Chaoxian Geng, Jon M. Babcock, and James D. Thomas. Sulfoxaﬂor and the sulfoximine insecticides: Chemistry, mode of action and basis for efﬁcacy on resistant insects. Pes- ticide Biochemistry and Physiology, 107(1):1–7, September 2013. ISSN 0048- 3575. doi: 10.1016/j.pestbp.2013.05.014. URL http://www.sciencedirect. com/science/article/pii/S0048357513000989.

[384] Alba Brancato, Daniela Brocca, Chloe De Lentdecker, Zoltan Erdos, Lucien Fer- reira, Luna Greco, Samira Jarrah, Dimitra Kardassi, Renata Leuschner, Christo- pher Lythgo, Paula Medina, Ileana Miron, Tunde Molnar, Alexandre Nougadere, Ragnor Pedersen, Hermine Reich, Angela Sacchi, Miguel Santos, Alois Stanek, Juergen Sturma, Jose Tarazona, Anne Theobald, Benedicte Vagenende, Alessia Verani, and Laura Villamar-Bouza. Modiﬁcation of the existing maximum residue levels for sulfoxaﬂor in grape leaves and similar species, and globe ar- tichokes. EFSA Journal, 15(11):e05070, November 2017. ISSN 1831-4732. doi: 10. 2903/j.efsa.2017.5070. URL https://efsa.onlinelibrary.wiley.com/doi/ abs/10.2903/j.efsa.2017.5070. 344 [385] Harry Siviter, Julia Koricheva, Mark J. F. Brown, and Ellouise Leadbeater. Quan- tifying the impact of pesticides on learning and memory in bees. Journal of Applied Ecology, 55(6):2812–2821, November 2018. ISSN 1365-2664. doi: 10. 1111/1365-2664.13193. URL https://besjournals.onlinelibrary.wiley. com/doi/abs/10.1111/1365-2664.13193.

[386] Zdenko Herceg and Thomas Vaissière. Epigenetic mechanisms and cancer: An interface between the environment and the genome. Epigenetics, 6(7):804–819, July 2011. ISSN 1559-2294. doi: 10.4161/epi.6.7.16262. URL https://doi.org/ 10.4161/epi.6.7.16262.

[387] Carmen J. Marsit. Inﬂuence of environmental exposure on human epigenetic regulation. The Journal of Experimental Biology, 218(Pt 1):71–79, January 2015. ISSN 1477-9145. doi: 10.1242/jeb.106971.

[388] Eric E. Nilsson, Ingrid Sadler-Riggleman, and Michael K. Skinner. Environ- mentally induced epigenetic transgenerational inheritance of disease. Environ- mental Epigenetics, 4(2), April 2018. doi: 10.1093/eep/dvy016. URL https: //academic.oup.com/eep/article/4/2/dvy016/5055600.

[389] Elizabeth J. Radford, Mitsuteru Ito, Hui Shi, Jennifer A. Corish, Kazuki Yamazawa, Elvira Isganaitis, Stefanie Seisenberger, Timothy A. Hore, Wolf Reik, Serap Erkek, Antoine H. F. M. Peters, Mary-Elizabeth Patti, and Anne C. Ferguson-Smith. In utero undernourishment perturbs the adult sperm methylome and intergenerational metabolism. Science, page 1255903, July 2014. ISSN 0036-8075, 1095- 9203. doi: 10.1126/science.1255903. URL http://science.sciencemag.org/ content/early/2014/07/09/science.1255903.

[390] Qi Chen, Wei Yan, and Enkui Duan. Epigenetic inheritance of acquired traits through sperm RNAs and sperm RNA modiﬁcations. Nature Reviews. Genetics, 17(12):733–743, 2016. ISSN 1471-0064. doi: 10.1038/nrg.2016.106.

345 [391] Wenchao Han, Ying Tian, and Xiaoming Shen. Human exposure to neonicotinoid insecticides and the evaluation of their potential toxicity: An overview. Chemosphere, 192:59–65, February 2018. ISSN 0045-6535. doi: 10.1016/j. chemosphere.2017.10.149. URL http://www.sciencedirect.com/science/ article/pii/S0045653517317332.

[392] Nobuhiko Hoshi, Tetsushi Hirano, Takuya Omotehara, Junko Tokumoto, Yuria Umemura, Youhei Mantani, Takashi Tanida, Katsuhiko Warita, Yoshiaki Tabuchi, Toshifumi Yokoyama, and Hiroshi Kitagawa. Insight into the mechanism of reproductive dysfunction caused by neonicotinoid pesticides. Biological & Phar- maceutical Bulletin, 37(9):1439–1443, 2014. ISSN 1347-5215.

[393] Verena Christen, Sara Bachofer, and Karl Fent. Binary mixtures of neonicotinoids show different transcriptional changes than single neonicotinoids in honeybees (Apis mellifera). Environmental Pollution, 220:1264–1270, January 2017. ISSN 0269-7491. doi: 10.1016/j.envpol.2016.10.105. URL http://www. sciencedirect.com/science/article/pii/S0269749116315482.

[394] Sipko van Dam, Urmo Võsa, Adriaan van der Graaf, Lude Franke, and João Pe- dro de Magalhães. Gene co-expression analysis for functional classiﬁcation and gene–disease predictions. Brieﬁngs in Bioinformatics, 19(4):575–592, July 2018. ISSN 1467-5463. doi: 10.1093/bib/bbw139. URL https://academic.oup.com/ bib/article/19/4/575/2888441.

[395] bumblebee - BioProject - NCBI. URL https://www.ncbi.nlm.nih.gov/ bioproject.

346