Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Elucidation of novel physiological and genetic elements associated with the cold adaptability and survival… Hingston, Patricia 2017

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

Item Metadata

Download

Media
24-ubc_2018_february_hingston_patricia.pdf [ 5.69MB ]
Metadata
JSON: 24-1.0362405.json
JSON-LD: 24-1.0362405-ld.json
RDF/XML (Pretty): 24-1.0362405-rdf.xml
RDF/JSON: 24-1.0362405-rdf.json
Turtle: 24-1.0362405-turtle.txt
N-Triples: 24-1.0362405-rdf-ntriples.txt
Original Record: 24-1.0362405-source.json
Full Text
24-1.0362405-fulltext.txt
Citation
24-1.0362405.ris

Full Text

i  ELUCIDATION OF NOVEL PHYSIOLOGICAL AND GENETIC ELEMENTS ASSOCIATED WITH THE COLD ADAPTABILITY AND SURVIVAL OF LISTERIA MONOCYTOGENES IN THE FOOD PROCESSING CONTINUUM by Patricia Hingston  B.A.Sc., Dalhousie University, 2010 M.Sc. Dalhousie University, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Food Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2017   © Patricia Hingston, 2017  ii  Abstract Novel physiological and genetic factors associated with the survival of Listeria monocytogenes in the food-processing continuum were investigated, with an emphasis on its cold-growth ability. Food-related L. monocytogenes strains (n=166) were sequenced and subsequently evaluated on their ability to tolerate cold (4°C), salt (6% NaCl, 25°C), acid (pH 5, 25°C), and desiccation (33% RH, 20°C) stress. Stress tolerances were associated with serotype, clonal complex, full-length inlA profiles, and plasmid harbourage. Notably, strains possessing full length inlA (as opposed to a truncated version) exhibited significantly (p<0.001) enhanced cold tolerance and plasmid-positive strains demonstrated enhanced (p=0.013) acid tolerance. Relative gene expression indicated that several plasmid-encoded genes (e.g., NADH peroxidase, clpL, proW) are induced in L. monocytogenes during growth in 6% NaCl and at pH 5. Additionally, a whole-genome sequence phylogeny revealed closely related stress sensitive and tolerant strains, highlighting that minor genetic differences impact strain phenotypes.  Strand-specific RNA sequencing showed that L. monocytogenes suppresses 1.3× more genes than it induces at 4°C relative to 20°C. The largest number (n=1,431) and greatest magnitude (>1,000-fold) of differentially expressed (e.g., >2-fold, p<0.05) genes occurred in late stationary-phase cells. A core set of 22 genes were upregulated at all five growth phases investigated and included nine genes required for branched-chain fatty acid (BCFA) synthesis. Correspondingly, BCFA levels increased by 15% during cold stress exposure. Transcription of antisense RNA (asRNA) was 2.5× higher in cells grown at 4°C relative to 20°C, with the most asRNA transcripts upregulated in lag phase cells. Spontaneous L. monocytogenes variants displaying enhanced cold tolerance (ECT) were isolated from a cold-sensitive strain culture following 84 days of storage at 4°C. While the parent iii  strain had an impaired ability to produce BCFAs, the ECT variants were able to overcome this limitation which is hypothesized to be a result of mutations identified in acetyl-coA carboxylase. Collectively this work has improved our understanding of the response of L. monocytogenes to to cold stress and genotypes associated with stress-tolerance phenotypes. This information may be useful for developing biomarkers to quickly predict the risks associated with food isolates, or aid in developing new and/or improved intervention strategies.      iv  Lay Summary Listeria monocytogenes is a potentially fatal human bacterial pathogen found ubiquitously in nature and accordingly, is frequently isolated from foods and food production facilities. While this pathogen is not heat tolerant, it can grow at temperatures as low as 0°C and therefore outbreaks are commonly associated with refrigerated, ready-to-eat foods that permit the pathogen to reach dangerous levels during the shelf-life. Currently, the regulations surrounding the presence of L. monocytogenes in foods assume that all strains behave similarly; however, we know this is not true. The goal of this research was to identify genetic elements associated with L. monocytogenes strains displaying enhanced abilities to tolerate food-related stresses (cold, salt, acid, and desiccation) with the hope of discovering biomarkers that can be used to quickly access the risks associated with a particular strain. Furthermore, this research aimed to improve our understanding of the ability of L. monocytogenes to grow at cold temperatures.      v  Preface A version of Chapter 2 has been published in the Frontiers in Microbiology Journal (Hingston, P., Chen, J., Dhillon, B. K., Laing, C., Bertelli, C., Gannon, V., Tasara, T., Allen, K., Brinkman, F. S., Hansen, L. T. and Wang, S. 2017. Genotypes associated with Listeria monocytogenes isolates displaying impaired or enhanced tolerances to cold, salt, acid, or desiccation stress. Front. Microbiol. 8:369 doi: 10.3389/fmicb.2017.00369). I was responsible for the majority of the work and manuscript preparation. Liang, C. assembled the whole genome sequences, Chen, J. conducted the targeted genomic element screenings, Bertelli, C. investigated the relationship between phenotypes and the presence of genomic islands, Dhillon, B. performed the single nucleotide polymorphism analysis and whole genome sequence phylogeny. Dr. Franco Pagotto and his team at the Listeriosis Reference Service at the Bureau of Microbial Hazards in Ottawa, Canada performed serotyping on our isolates. Wang, S., Truelstrup Hansen, L., and Brinkman, F. and Chen, J. helped with reviewing the experimental design and reviewing and editing the manuscript. Allen K. was involved in the concept formation and Gannon, V. and Tasara, T. generously donated their L. monocytogenes strains and contributed to the final editing of the manuscript. Wang, S. was the supervisory author.  I performed and analyzed the research described in Chapters 3 and 5. Chen, J., Truelstrup Hansen, L., and Wang, S. provided feedback on the experimental design and helped with critical review and editing of the chapters.  A version of Chapter 4 has been published in the PLoS ONE journal (Hingston, P., Chen, J., Allen, K., Hansen, L. T., and Wang, S. 2017. Strand specific RNA-sequencing and membrane lipid profiling reveals growth phase-dependent cold-stress response mechanisms in Listeria monocytogenes. PLoS ONE, 12(6), e0180123). I was responsible for the majority of the work and vi  manuscript preparation. Chen, J., Wang, S., and Truelstrup Hansen, L. helped in reviewing the experimental design and reviewing and editing the manuscript.  Allen K. was involved in the concept formation. Wang, S. was the supervisory author.    vii  Table of Contents  Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ........................................................................................................................ vii List of Tables ................................................................................................................................xv List of Figures ............................................................................................................................ xvii List of Supplementary Materials ................................................................................................xx List of Symbols ........................................................................................................................... xxi List of Abbreviations ................................................................................................................ xxii Acknowledgements ....................................................................................................................xxv Dedication ................................................................................................................................ xxvii Chapter 1: Introduction and literature review ...........................................................................1 1.1 Introduction ................................................................................................................... 1 1.2 Research overview ........................................................................................................ 4 1.3 Literature review ........................................................................................................... 6 1.3.1 Overview of Listeria monocytogenes ................................................................. 6 1.3.1.1 L. monocytogenes genomics ............................................................................... 9 1.3.1.2 L. monocytogenes subtypes ............................................................................... 11 1.3.1.2.1 Genetic lineages and serotypes ................................................................... 11 1.3.1.2.2 Clonal complexes and sequence types ........................................................ 13 1.3.1.3 L. monocytogenes pathogenicity ....................................................................... 14 viii  1.3.2 L. monocytogenes physiology and food-related stress tolerances .................... 18 1.3.2.1 Acid tolerance ................................................................................................... 20 1.3.2.2 Cold and heat tolerance ..................................................................................... 21 1.3.2.3 Osmotolerance .................................................................................................. 24 1.3.2.4 Desiccation tolerance ........................................................................................ 26 1.3.3 Factors associated with L. monocytogenes phenotypes important to the food industry…………………….. ........................................................................... 29 1.3.3.1 Strain origin ...................................................................................................... 29 1.3.3.2 L. monocytogenes subtypes ............................................................................... 31 1.3.3.3 Persistent vs. sporadic strains ........................................................................... 33 1.3.3.4 Presence or absence of genes ............................................................................ 35 1.3.3.5 Differential expression of genes ....................................................................... 37 1.3.3.6 Horizontal gene transfer .................................................................................... 39 1.3.3.6.1 Plasmids ...................................................................................................... 40 1.3.3.6.2 Genomic islands .......................................................................................... 42 1.3.3.6.3 Bacteriophages ............................................................................................ 45 1.3.3.6.4 Transposons ................................................................................................. 46 1.3.3.7 Single nucleotide polymorphisms (SNPs) ........................................................ 47 1.3.3.8 Regulatory RNAs .............................................................................................. 50 1.3.4 L. monocytogenes stress response mechanisms ............................................... 50 1.3.4.1 Membrane lipid compositional changes ........................................................... 51 1.3.4.2 Regulatory elements.......................................................................................... 53 1.3.4.2.1 RNA polymerase sigma factors................................................................... 53 ix  1.3.4.2.2 Global regulatory proteins ........................................................................... 54 1.3.4.2.3 Two-component systems ............................................................................. 54 1.3.4.2.4 Non-coding RNA ........................................................................................ 55 1.3.4.3 Stress induced proteins ..................................................................................... 56 1.3.4.3.1 Osmolyte and oligopeptide transporters ...................................................... 56 1.3.4.3.2 Cold shock proteins ..................................................................................... 57 1.3.4.3.3 Heat shock proteins ..................................................................................... 57 1.3.4.3.4 RNA and DNA repair proteins .................................................................... 58 Chapter 2: Genotypes associated with Listeria monocytogenes isolates displaying impaired or enhanced tolerances to cold, salt, acid, or desiccation stress ..............................................59 2.1 Introduction ................................................................................................................. 59 2.2 Materials and methods ................................................................................................ 62 2.2.1 Isolates and culture conditions ......................................................................... 62 2.2.2 Whole genome sequencing ............................................................................... 62 2.2.3 Lineage determination ...................................................................................... 63 2.2.4 Multi locus sequence typing ............................................................................. 63 2.2.5 In silico serogroup/serotype assignment .......................................................... 63 2.2.6 Targeted genomic element screenings ............................................................. 64 2.2.7 Identification of putative plasmid contigs ........................................................ 64 2.2.8 Cold tolerance assay ......................................................................................... 65 2.2.9 Salt and acid tolerance assay ............................................................................ 65 2.2.10 Desiccation tolerance assay .............................................................................. 66 2.2.11 Phenotype designations and statistical analyses ............................................... 66 x  2.2.12 Phylogenetic reconstruction based on core genome single nucleotide variants 68 2.2.13 SNP detection ................................................................................................... 68 2.2.14 Statistical methods for elucidating SNPs associated with stress-tolerance phenotypes ........................................................................................................ 69 2.2.15 Genomic islands analysis ................................................................................. 69 2.3 Results ......................................................................................................................... 70 2.3.1 Genetic characteristics of L. monocytogenes isolates based on WGS data ...... 70 2.3.2 Stress tolerance distributions among L. monocytogenes isolates ..................... 75 2.3.3 Overlapping stress-tolerance phenotypes ......................................................... 76 2.3.4 Stress tolerances of L. monocytogenes lineages, serotypes, and clonal complexes ......................................................................................................... 77 2.3.5 Associations between plasmid harbourage and stress tolerances ..................... 81 2.3.6 Associations between inlA profiles and stress tolerances ................................ 82 2.3.7 Associations between stress tolerances and the presence of SSI1 or LGI1 ..... 83 2.3.8 SNP analyses of stress-sensitive and tolerant isolates ...................................... 83 2.3.9 Genomic islands of stress-sensitive and tolerant isolates ................................. 86 2.4 Discussion ................................................................................................................... 86 2.4.1 The tolerance of L. monocytogenes to food-related stresses differs between and within lineages, serotypes, and clonal complexes ............................................ 86 2.4.1.1 Cold stress ......................................................................................................... 86 2.4.1.2 Salt stress .......................................................................................................... 88 2.4.1.3 Acid stress ......................................................................................................... 88 2.4.1.4 Desiccation stress .............................................................................................. 89 xi  2.4.2 Certain genetic elements are associated with the stress tolerance of L. monocytogenes ................................................................................................. 90 2.4.2.1 Plasmids ............................................................................................................ 90 2.4.2.2 Full length inlA ................................................................................................. 93 2.4.2.3 SSI-1 ................................................................................................................. 94 2.4.2.4 LGI1 .................................................................................................................. 94 2.4.2.5 SNPs associated with stress-tolerance phenotypes ........................................... 95 2.4.2.6 Genomic islands ................................................................................................ 97 2.5 Conclusions ................................................................................................................. 98 Chapter 3: Comparative analysis of Listeria monocytogenes plasmids and expression levels of plasmid-encoded genes during growth under salt and acid stress conditions .................100 3.1 Introduction ............................................................................................................... 100 3.2 Materials and methods .............................................................................................. 102 3.2.1 Strains and culture conditions ........................................................................ 102 3.2.2 Genetic characterization of plasmids ............................................................. 103 3.2.3 RNA isolation and real-time qPCR analysis .................................................. 105 3.2.4 Stress tolerance comparisons of wildtype and plasmid-cured strains ............ 106 3.3 Results and discussion .............................................................................................. 107 3.3.1 Plasmid types and characteristics ................................................................... 107 3.3.2 Genetic elements shared by group 1 and group 2 plasmids ........................... 112 3.3.3 Genetic elements associated with group 1 plasmids ...................................... 120 3.3.4 Genetic elements associated with group 2 plasmids ...................................... 124 xii  3.3.5 Expression of L. monocytogenes plasmid encoded genes during growth under salt and acid stress conditions ........................................................................ 128 3.3.6 Stress tolerance comparisons of wildtype and plasmid-cured strains ............ 134 3.4 Conclusions ............................................................................................................... 137 Chapter 4: Strand specific RNA-sequencing and membrane lipid profiling reveals growth-phase dependent cold-stress response mechanisms in Listeria monocytogenes ....................139 4.1 Introduction ............................................................................................................... 139 4.2 Materials and methods .............................................................................................. 143 4.2.1 Culture conditions and time point selection ................................................... 143 4.2.2 Fatty acid analysis .......................................................................................... 144 4.2.3 RNA isolation and sequencing ....................................................................... 144 4.2.4 RNA-seq data analysis ................................................................................... 145 4.2.5 Clustering of gene expression profiles ........................................................... 146 4.2.6 Functional categorization of differentially expressed genes .......................... 146 4.2.7 Quantitative PCR validation of RNA-seq data ............................................... 147 4.2.8 Accession numbers ......................................................................................... 148 4.3 Results and discussion .............................................................................................. 148 4.3.1 mRNA transcriptome of the L. monocytogenes cold-stress response ............ 148 4.3.2 Core set of genes induced by cold stress ........................................................ 152 4.3.3 Core set of genes suppressed under cold stress .............................................. 160 4.3.4 L. monocytogenes cold-stress response at individual growth phases ............. 162 4.3.4.1 Early lag phase (G1) ....................................................................................... 162 4.3.4.2 Transition to exponential growth phase (G2) ................................................. 167 xiii  4.3.4.3 Mid-exponential growth phase (G3) ............................................................... 168 4.3.4.4 Transition to stationary phase (G4)................................................................. 168 4.3.4.5 Late-stationary phase (G5) .............................................................................. 169 4.3.5 L. monocytogenes cold stress regulon: A comparison with previous studies 170 4.3.5.1 Osmolyte and oligopeptide uptake.................................................................. 171 4.3.5.2 RNA and DNA repair ..................................................................................... 174 4.3.5.3 Regulatory elements........................................................................................ 176 4.3.5.4 Ribosome functions ........................................................................................ 182 4.3.5.5 Cold-stress proteins ......................................................................................... 183 4.3.5.6 Additional proteins with putative roles in the L. monocytogenes cold-stress response .......................................................................................................... 183 4.3.6 Cold-induced membrane lipid composition changes ..................................... 185 4.3.6.1 Increase in anteiso C15:0 ................................................................................ 186 4.3.6.2 Shortening of fatty acid chain lengths ............................................................ 188 4.3.6.3 Increase in unsaturated fatty acids .................................................................. 189 4.3.7 Comparison of antisense transcription at 20°C and 4°C ................................ 192 4.4 Conclusions ............................................................................................................... 198 Chapter 5: Phenotypic and genetic characterization of Listeria monocytogenes enhanced cold tolerance variants isolated from prolonged cold storage cultures ................................200 5.1 Introduction ............................................................................................................... 200 5.2 Materials and methods .............................................................................................. 202 5.2.1 Strains and culture conditions ........................................................................ 202 5.2.2 Long-term cold stress exposure ...................................................................... 202 xiv  5.2.3 Screening for variants with enhanced cold tolerance ..................................... 202 5.2.4 Stress tolerance profiling ................................................................................ 203 5.2.5 Cell morphology and colony enumeration ..................................................... 205 5.2.6 Membrane lipid profiling ............................................................................... 205 5.2.7 Whole genome sequencing and SNP analysis ................................................ 206 5.3 Results and discussion .............................................................................................. 207 5.3.1 Long-term cold stress survival ....................................................................... 207 5.3.2 L. monocytogenes variants with enhanced cold tolerance (ECT) .................. 212 5.3.2.1 Membrane lipid profiles .................................................................................. 215 5.3.2.2 Stress tolerance profiling ................................................................................ 217 5.3.2.3 SNPs detected in ECT variants ....................................................................... 221 5.4 Conclusions ............................................................................................................... 228 Chapter 6: Conclusions .............................................................................................................230 6.1 Future directions ....................................................................................................... 235 References ...................................................................................................................................237 Appendices ..................................................................................................................................274 Appendix A - Chapter 2 supplementary figure ....................................................................... 274 Appendix B - Chapter 3 supplementary table ......................................................................... 275 Appendix C - Chapter 4 supplementary figure ....................................................................... 276 Appendix D - Chapter 5 supplementary tables ....................................................................... 277  xv  List of Tables Table 1-1. Characteristics of 24 listeriosis outbreaks by implicated food categories, Foodborne Disease Outbreak Surveillance System, United States, 1998-2008. ............................................... 9 Table 1-2. Summary of L. monocytogenes lineages.. ................................................................... 12 Table 2-1. Genetic characteristics and prevalence of sensitive and tolerant phenotypes among L. monocytogenes belonging to different serotypes. ......................................................................... 73 Table 2-2. Genetic characteristics and prevalence of sensitive and tolerant phenotypes among L. monocytogenes belonging to different clonal complexes. ............................................................ 74 Table 3-1. L. monocytogenes strains used in this study. ............................................................ 103 Table 3-2. Characteristics of G1 plasmids and their associated strains. .................................... 109 Table 3-3. Characteristics of G2 plasmids and their associated strains. .................................... 110 Table 3-4. Predicted proteins observed on both group 1 and group 2 plasmids.. ...................... 116 Table 3-5. Predicted proteins uniquely observed on group 1 plasmids. .................................... 122 Table 3-6. Predicted proteins uniquely observed on group 2 plasmids.. ................................... 127 Table 4-1. Primers used for quantitative PCR validation of RNA-seq data. ............................. 147 Table 4-2. Core set of genes upregulated  across multiple growth phases in L. monocytogenes cells at 4°C vs. 20°C. .................................................................................................................. 155 Table 4-3. Pathways and gene ontology processes significantly enriched among genes upregulated at 4°C vs. 20°C. ....................................................................................................... 158 Table 4-4. Pathways and gene ontology processes significantly  enriched among genes downregulated at 4°C vs. 20°C. .................................................................................................. 162 Table 4-5. Top 10 most highly induced genes in L. monocytogenes cells at 4°C vs 20°C at five different growth phases. .............................................................................................................. 165 xvi  Table 4-6. Genes commonly associated with the L. monocytogenes CSR. ............................... 172 Table 4-7. Transcription regulators significantly overrepresented among genes differentially expressed at 4°C vs. 20°C. .......................................................................................................... 181 Table 4-8. Top 20 most highly expressed antisense transcripts in L. monocytogenes at 20°C and 4°C. ............................................................................................................................................. 195 Table 5-1. Parent strain and select enhanced cold tolerant variants selected for sequencing and additional experimentation.......................................................................................................... 207 Table 5-2. Single nucleotide polymorphisms identified in L. monocytogenes enhanced cold tolerance variants. ....................................................................................................................... 227  xvii  List of Figures Figure 1-1. Distribution of L. monocytogenes lineages among different ecological compartments.. .............................................................................................................................. 12 Figure 1-2. Schematic representation of the physiopathology of L. monocytogenes infection. .. 15 Figure 1-3. Specificity of L. monocytogenes invasion proteins internalin A (InlA) and B (InlB) for human, guinea pig, and mouse and rat Met and E-cadherin intestinal cell receptors. ............ 16 Figure 2-1. Stress tolerance distributions of 166 L. monocytogenes isolates. ............................. 76 Figure 2-2. Numbers of L. monocytogenes isolates with multiple sensitivities or tolerances to food-related stresses.. .................................................................................................................... 77 Figure 2-3. Levels of tolerance to food-related stresses among L. monocytogenes serotypes.. .. 80 Figure 2-4. Levels of tolerance to food-related stresses of different L. monocytogenes clonal complexes.. ................................................................................................................................... 81 Figure 2-5. Prevalence of full length inlA and plasmid harbourage among L. monocytogenes stress-tolerance phenotypes.. ........................................................................................................ 82 Figure 2-6. Whole genome SNP phylogeny of 166 L. monocytogenes isolates and their associated genetic characteristics and stress-tolerance phenotypes. ............................................. 85 Figure 3-1. Neighbouring Joining phylogenic tree showing the genetic groupings of L. monocytogenes G1 and G2 plasmids. ......................................................................................... 111 Figure 3-2. Complete alignment of L. monocytogenes G1 plasmid sequences.. ....................... 118 Figure 3-3. Complete alignment of L. monocytogenes G1 plasmid sequences. ........................ 119 Figure 3-4. Differential expression of L. monocytogenes plasmid-encoded genes in strains subjected to salt and acid stress in comparison to the control. ................................................... 133 xviii  Figure 3-5. Growth of wildtype L. monocytogenes strains and their plasmid-cured (PC) counterparts in BHIB, BHIB +6% NaCl, and BHIB pH 5 at 25°C.. .......................................... 136 Figure 4-1. Cell growth phase sampling strategy at 4°C and 20°C. .......................................... 144 Figure 4-2. Principle component analysis plot of RNA sequencing biological replicates.. ...... 150 Figure 4-3. Numbers of L. monocytogenes sense and antisense RNAs differentially expressed at five growth phases in response to cold stress.. ........................................................................... 151 Figure 4-4. Heatmaps showing the number of L. monocytogenes sense and antisense RNA co-upregulated or co-downregulated between pairs of growth phases at 4°C. ................................ 152 Figure 4-5. Differential gene expression patterns observed in L. monocytogenes cells grown across five growth phases at 4°C. ............................................................................................... 159 Figure 4-6. Relative proportions of specific branched-chain fatty acids out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4°C and 20°C. ......................... 188 Figure 4-7. Relative proportion of short chain and unsaturated fatty acids out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4 and 20°C.. ........................ 189 Figure 4-8. Proportions of anteiso C15:0 and the combined sum of the unsaturated fatty acids 16:1 cis-Δ9 and 18:1 cis-Δ9 out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4°C. .......................................................................................................... 192 Figure 4-9. Coverage maps of select highly expressed antisense transcripts in L. monocytogenes cells grown at 20 or 4°C. ............................................................................................................ 197 Figure 5-1. Growth and survival of various L. monocytogenes strains throughout one year of storage at 4°C in BHIB. .............................................................................................................. 211 Figure 5-2. Growth of Lm96 and three of its ECT variants at 4°C in BHIB.. ........................... 214 xix  Figure 5-3. Colony morphology of Lm96, and an Lm96 ECT variant with a pinpoint colony morphology, and cellular morphology of Lm96 and all ECT variants. ...................................... 214 Figure 5-4. Relative proportions of anteiso C15:0, and C16:1 cis-Δ9 and C18:1 cis- Δ9 in the membranes of exponential cells of Lm96 and three ECT variants grown at 20 or 4°C in BHIB ...................................................................................................................................................... 217 Figure 5-5. Growth or survival comparisons of Lm96 and three enhanced cold tolerance variants under cold, salt, acid, heat, and desiccation stress.. .................................................................... 220 Figure 5-6. Molecular clock based whole genome sequence phylogeny of Lm96 and three of its ECT variants.. ............................................................................................................................. 228  xx  List of Supplementary Materials Supplementary Table 2-1. Summary of all L. monocytogenes isolates and their associated genetic and phenotypic characteristics. Supplementary Table 2-2. SNVs uniquely identified among cold sensitive isolates. Supplementary Table 2-3. SNVs uniquely identified among salt sensitive isolates. Supplementary Table 2-4. SNVs uniquely identified among acid sensitive isolates Supplementary Table 2-5. SNVs uniquely identified among desiccation sensitive isolates. Supplementary Table 2-6. SNVs uniquely identified among cold tolerant isolates. Supplementary Table 2-7. SNVs uniquely identified among salt tolerant isolates. Supplementary Table 2-8. SNVs uniquely identified among acid tolerant isolates. Supplementary Table 2-9. SNVs uniquely identified among desiccation tolerant isolates. Supplementary Table 4-1. Differential expression and associated cluster memberships of L. monocytogenes EGD ORFs and antisense RNA at 4°C relative to 20°C. Supplementary Table 4-2. Normalized paired-end read counts for all L. monocytogenes EGD ORFs and antisense RNAs at 4 and 20°C. Supplementary Table 4-3. Fatty acid membrane profiles of L. monocytogenes at 4°C and 20°C.    xxi  List of Symbols aw Water activity h Hour(s) Nmax Maximum cell density s Second(s) µmax Maximum growth rate v/v Volume per volume w/v Weight per volume  xxii  List of Abbreviations A absorbance aa Amino acid a-Cx:0 Anteiso-fatty acid with x carbons ACP Acyl-carrier protein aw Water activity AS Acid sensitive asRNA Antisense RNA AT Acid tolerant AT Adenine + thymine  ATP Adenosine triphosphate  bp Base pair(s) BCAA Branched-chain amino acid BCFA Branched-chain fatty acid BHIB Brain heart infusion broth CAP Cold acclimation protein CC Clonal complex CDC Centre for disease control CFU Colony forming units CIP Cold induced proteins CS Cold sensitive CSP Cold stress proteins CSR Cold stress response CT Cold tolerant DE Differential expression/expressed DNA Deoxyribonucleic acid DS Desiccation sensitive DT Desiccation tolerant ECT Enhanced cold tolerance EPS Extracellular polymeric substance(s) EU European Union FA Fatty acid FAME Fatty acid methyl ester FDA Food and Drug Administration G1, G2 Plasmid groups 1 and 2 (based on repA sequence) G1-G5 Growth phases 1 (lag phase) through 5 (late-stationary phase) GASP Growth advantage in stationary phase GC Guanine + cytosine GO Gene ontology GWAS Genome-wide association study(ies) HCl Hydrochloric acid HK Histidine kinase HSP Heat shock proteins xxiii  i-Cx:0 Iso fatty acid with x carbons kbp Kilobase pair(s) LGI-1 Listeria genomic island 1 LGI-2 Listeria genomic island 2 LI Lineage I LII Lineage II LIII Lineage III lin Genes from Listeria innocua LIPI-1 Listeria pathogenicity island 1 LIPI-2 Listeria pathogenicity island 2 LIPI-3 Listeria pathogenicity island 3 lmo Genes from Listeria monocytogenes EGD-e LMON Genes from Listeria monocytogenes EGD LPD Lag phase duration M molar MATE Multi-antimicrobial extrusion protein Mbp Mega base pairs MCO Multicopper oxidase MDR Multi-drug resistance MGE Mobile genetic element MIC Minimum inhibitory concentration ml Milliliter  MLST Multi locus sequence typing  MSE Mean standard error NaCl Sodium chloride NCBI National Centre for Biotechnology Information ncRNA Non-coding RNA Nmax Maximum cell density nt Nucleotide(s) ORF Open reading frame PC Plasmid cured PCA Principle component analysis PBS Phosphate buffered saline PCR Polymerase chain reaction PDC Pyruvate dehydrogenase complex PE Paired-end PFGE Pulsed-field gel electrophoresis PMN Polymorphonuclear neutrophil PMSC Premature stop codon PS Peptone saline PTS Phosphotransferase system qPCR Quantitative PCR RH Relative humidity RIN RNA integrity number RNA Ribonucleic acid xxiv  RR Response regulator rRNA Ribosomal RNA RTE Ready-to-eat  SCFA Straight-chain fatty acids SD Standard deviation SFA Saturated fatty acid SMR Small multi-drug resistance SNP Single nucleotide polymorphism sRNA Small RNA SS Salt sensitive SSI-1 Stress survival islet 1 SSI-2 Stress survival islet 2 ST Sequence type ST Salt tolerant std Standardized value TCS Two-component system TSA Tryptic soy agar TSB Tryptic soy broth TRG Time to detectable re-growth tRNA Transfer RNA TSB Tryptic soy broth UFA Unsaturated fatty acids US United States UTR Untranslated region WGS Whole genome sequence/sequencing YE Yeast extract   xxv  Acknowledgements First, I would like to express my gratitude to Dr. Kevin Allen whose vision for food safety inspired me to move across the country for my PhD. Though we did not get to work together as much as we had hoped, his enthusiasm for science and strive for excellence has made me a better researcher and communicator.    Secondly, I express my sincere appreciation for Dr. Lisbeth Truelstrup Hansen who was the first to introduce me to food microbiology but has since become a life-long friend. Lisbeth has been a true mentor to me for 10 years and has always gone above and beyond her responsibilities as first my Undergraduate and Masters’ supervisor and later as a member of my PhD committee. Her continuous support and encouragement has been instrumental in the completion of this thesis.  I would also like to extend a very heartfelt thankyou to Dr. Siyun Wang who graciously welcomed me as one of her students. Siyun’s positive outlook on life and dedication to her students is both admirable and contagious. Her ongoing encouragement and confidence in my abilities has helped me to achieve new milestones in my academic career.    It is also imperative that I address my appreciation for Dr. Jessica Chen, an amazing post doc and an even more remarkable friend. Jess’ wealth of knowledge is only matched by her desire to share it with others. It was truly and honour and an irreplaceable experience to work alongside her during my PhD.  Gratitude also goes out to Dr. Chad Leung, Dr. Victor Gannon, and Dr. Franco Pagotto for their collaboration and willingness to help in times of need. I’d also like to thank Dr. Gannon and Dr. Taurai Tasara for generously donating the Listeria monocytogenes isolates for this work.  I would also like to thank my committee members Drs. Rickey Yada, Fiona Brinkman, and Brett Finlay for their insightful contributions and encouragement throughout my PhD.  xxvi  I’d also like to acknowledge the financial support for this study and my PhD which came from the Four Year Doctoral Fellowship from the University of British Columbia, the National Sciences and Engineering Research Council of Canada, and Alberta Innovates BioSolutions.   A special thank you to my fellow lab members both past and present for the many walks to get caffeine, and for enriching my life here in Vancouver. Specifically, I’d like to thank Karen Fong for her endearing friendship and cheerful attitude which made working weekends something to look forward to, and Justin Falardeau for always offering a helping a hand (or hug) at the right time.  Finally, I would like to extend my sincere thanks to my family and friends who always believed in me and who took the time to listen to me describe the research that inspired me, even if it was confusing. I especially thank my husband Maksym Opushnyev for his never-ending technical and mechanical support, for biking home with me late at night, and for making each day end on a good note.  xxvii  Dedication I dedicate this work to my loving husband whom were it not for this PhD, I would have never met. This journey has been as much mine as it has been ours.     1  Chapter 1: Introduction and literature review 1.1 Introduction  In recent years, foods contaminated with the foodborne bacterial pathogen Listeria monocytogenes have been a significant issue in North American and internationally, causing repeated recalls and outbreaks of foodborne illness. Many of these outbreaks have been attributed to the bacterium’s ability to survive various stresses employed by the food industry to eliminate or suppress bacterial growth and survival in food processing environments and final products. One of the most unique characteristics contributing to the success of L. monocytogenes as a foodborne pathogen is its ability to grow at temperatures as low as -0.4°C (Walker et al., 1990), rendering refrigeration an ineffective post-processing control. Levels of L. monocytogenes in foods following production are usually low and unlikely to cause disease, it is therefore refrigerated, ready-to-eat (RTE) foods with extended shelf lives that present the largest risk to consumers. Significant ingestion of foods contaminated with L. monocytogenes can cause listeriosis, a very serious disease with high mortality rates of 20-40% (Swaminathan and Gerner-Smidt, 2007). Populations that are more susceptible to listeriosis include immunocompromised individuals, the elderly, pregnant women, and unborn or newborn babies (Farber and Peterkin, 1991). Both Canada and the EU have adopted regulations for the control of L. monocytogenes in RTE-foods (Health Canada, 2016; Luber, 2011), allowing the presence of up to 100 CFU/g in foods that do not permit growth beyond this level within the shelf-life of the product, and a zero tolerance policy for foods identified as supporting growth. In the US, the zero tolerance policy is applied to all food products (US FDA, 2015). However, despite these regulations and industry awareness, widespread outbreaks continue to occur. So far in 2017, one outbreak has been linked  2  to the consumption of soft raw-milk cheese in the US that resulted in eight hospitalizations and two deaths (US CDC, 2017). Previously in 2016, a total of 20 hospitalizations and five deaths occurred from three listeriosis outbreaks implicating frozen vegetables, packaged salads, and raw milk (US CDC, 2017) with the packaged salads outbreak impacting Canadians as well. These numbers did not exceed those of 2015 where two multistate outbreaks involving soft cheese and ice cream resulted in 40 illnesses and six deaths (US CDC, 2017). Collectively it is estimated that approximately 2500 cases of clinical human listeriosis (including 500 deaths) occur each year in the US, making it the third leading cause of death from foodborne illnesses (Mead et al., 1999; Scallan et al., 2011).  In addition to the large economic burden that recalls (Ivanek et al., 2005), and outbreaks (Scharff, 2012) place on society, hundreds of non-outbreak associated products equating to millions of pounds of food are recalled each year for the presumptive presence of L. monocytogenes (Leonhardt, 2016). The increasing consumer demand for fresh, minimally processed RTE-foods combined with the ever-growing global population, necessitates the need for improved control of L. monocytogenes in food production environments or alternatively, a more comprehensive understanding of the risks associated with the presence of specific strains.  It is well known for L. monocytogenes and bacterial species at large, that strains vary in their abilities to tolerate different stresses and for pathogenic bacteria, this extends to their ability to cause disease. Based on this, some strains pose a larger concern to consumer health than others. If we could develop a tool to quickly identify the risks associated with a specific isolate, we could potentially avoid future recalls and reduce our levels of food waste. The most efficient way to do this is through the use of biomarkers which in this instance refers to a genetic feature that is known to be associated with a particular phenotype. However, to date very few genotype-phenotype  3  associations have been uncovered in L. monocytogenes. This is in part due to the fact that the tools needed (e.g., whole genome sequencing) have only recently been made widely available, and also the complexity of analyzing the genetic differences between strains makes it difficult to identify a single gene or mutation responsible for a specific phenotype. As a result, current knowledge has predominantly focused on larger associations such as those between phenotypes and L. monocytogenes sources of isolation, lineages, and serotypes.  Another concern surrounding the current Canadian and EU regulations for the presence and control of L. monocytogenes in foods is how we validate the potential for L. monocytogenes growth in foods. Health Canada requires producers of refrigerated RTE-foods to perform challenge testing on their products where the growth potential of a cocktail of three to five L. monocytogenes strains is evaluated throughout the product shelf-life (Health Canada, 2012). While it is suggested to include outbreak strains in this cocktail as well as strains that have been carefully characterized, it is unlikely that these five strains represent the extremes of the survival capabilities of L. monocytogenes. Furthermore, bacteria are constantly evolving and acquiring mutations that can improve their survival in their surrounding environment, making it difficult to estimate the growth potential of all L. monocytogenes biotypes. The survival capabilities of the strains encountered in the food industry today will likely differ from those isolated a decade from now and so on. At this time, little is known regarding how food safety control measures may be selecting for the evolution of L. monocytogenes strains with enhanced abilities to tolerate food-related stresses. More research is needed in this area to assess the potential long-term impacts of current intervention strategies.   4  1.2 Research overview The encompassing objective of this thesis was to elucidate novel physiological and genetic factors associated with the ability of L. monocytogenes strains to survive in the food-processing continuum with a particular focus on the distinctive ability of L. monocytogenes to tolerate cold stress. To accomplish this, the research for this thesis was divided into four parts based on the following four hypotheses.  Hypothesis 1: L. monocytogenes strains with similar stress-tolerance phenotypes share common genetic elements that can serve as biomarkers for quickly predicting isolate phenotypes and the potential risks that they may pose.  Research objective: A large collection of 166 L. monocytogenes food isolates will be sequenced and subsequently evaluated on their ability to tolerate cold (4°C), salt (6% NaCl), acid (pH 5), and desiccation (33% RH) stress. Statistical and bioinformatics analyses will then be used to elucidate potential genotype-phenotype associations.  Hypothesis 2: The L. monocytogenes cold-stress response is growth-phase dependent and facilitated by non-coding, antisense RNA.  Research objective: Strand-specific RNA sequencing and membrane lipid profiling will be conducted on a fast cold-growing strain of L. monocytogenes at five distinct growth phases following a downshift in temperature from 20 to 4°C. The L. monocytogenes cold-stress response at each of the growth phases will then be analyzed and compared.  5  Hypothesis 3: L. monocytogenes plasmids contain genes that contribute to the bacterium’s ability to tolerate salt and acid stress.   Research objective: • Genes found on L. monocytogenes plasmids will be analyzed and the expression levels of prevalent stress response genes will be compared in exponential phase cells grown under control [brain heart infusion broth (BHIB), 30°C], salt (BHIB+6% NaCl, 30°C), and acid stress (BHIB pH 5, 30°C) conditions using quantitative real-time PCR.   Hypothesis 4: Extended cold stress exposure promotes the formation of L. monocytogenes variants with enhanced tolerance to cold and other food-related stresses. Research objective: • A selection of L. monocytogenes strains will be stored in BHIB at 4°C for one year and monthly aliquots will be screened for the presence of variants exhibiting enhanced cold tolerance relative to the parent strains. Isolated variants will then be tested for enhanced tolerance to other food related stresses (salt, acid, desiccation, heat). Whole genome sequencing and membrane lipid profiling will be used to determine if any physiological or genetic differences exist between any variant isolate and their parent strains. It should be noted that within this thesis the term isolate is used to denote L. monocytogenes cells that have been isolated from an environment but not yet sequenced or thoroughly characterized. On the other hand, the term strain is used to denote L. monocytogenes isolates that have been sequenced and/or characterized thoroughly enough to have confidence that they represent a unique isolate within a given collection.   6  1.3 Literature review 1.3.1 Overview of Listeria monocytogenes Listeria monocytogenes belongs to the Listeria genus which contains Gram-positive, rod shaped, non-spore forming and facultative anaerobic bacteria that have peritrichous flagella at temperatures below 37°C. The Listeria genus currently consists of 17 species including L. grayi, L. innocua, L. ivanovii, L. monocytogenes, L. seeligeri, and L. welshimeri, as well as nine other recently identified species: L. aquatica, L. cornellensis, L. fleischmannii, L. floridensis, L. grandensis, L. marthii, L. riparia, L. rocourtiae, L. newyorkensis, L. booriae, and L. weihenstephanesis [reviewed in Orsi and Wiedmann, (2016)]. Taxonomically, the Listeria genus is closely related to the genera Brochothrix, Bacillus, Lactobacillus, and more distantly to Streptococcus, Lactococcus, Enterococcus and Staphylococcus (Farber and Peterkin, 1991). Among Listeria spp. only two are considered pathogenic: L. monocytogenes and L. ivanovii with L. ivanovii predominantly affecting animals though in a few rare cases it has been associated with human listeriosis (Cummins et al., 1994; Guillet et al., 2010; Snapir et al., 2006). L. monocytogenes on the other hand is known to commonly infect both humans and animals and notably was also the first Listeria spp. discovered. In 1926, L. monocytogenes was discovered to be the causative agent of an epidemic affecting rabbits and guinea pigs in England (Murray et al., 1926). Human cases of listeriosis began to be reported as early as 1929. However, the bacterium was not recognized as a serious human pathogen until the first major foodborne outbreak occurred in Nova Scotia, Canada in 1981 where the consumption of contaminated coleslaw in a hospital led to 41 cases of listeriosis and 18 deaths in pregnant women and neonates (Schlech and Acheson, 2000). Since then, L. monocytogenes has been widely acknowledged as a very serious pathogen  7  often causing life-threatening foodborne disease in susceptible persons which includes infants, pregnant women, and elderly and immunocompromised individuals (Lorber, 2005). The Listeria genus can be divided into two groups based on the relatedness of the species to L. monocytogenes. The first group, named Listeria sensu strictu includes L. monocytogenes, L. seeligeri, L. ivanovii, L. welshirmeri, L. innocua and L. marthii while the second group named Listeria sensu lato encompasses all other species (Chiara et al., 2015; Orsi and Wiedmann, 2016). It is well documented that species within Listeria sensu strictu are widely distributed in the environment throughout the world with soil, water, and animals being the most common reservoirs (Gelbíčová and Karpíšková, 2012). Listeria spp. are considered saprophytic organisms meaning they obtain nutrients from dead or decomposing matter (Weis and Seeliger, 1975; Welshimer, 1960), thus allowing them to survive for long periods of time in plant-soil environments. In a study in Austria (Linke et al., 2014) 149 of 467 soil samples (30%) were positive for Listeria spp. with L. seeligeri being the most prevalent (15% of samples), followed by L. monocytogenes and L. innocua (6% of samples). In a study in Canada Listeria spp. were isolated from 35% and 72% of fresh water samples from an urban and a rural watershed, respectively (Stea et al., 2015). Amongst animals, incidences of L. monocytogenes carriage have been reported as high as 91.5% among chickens sampled from Nigeria (Ishola et al., 2016), and in a study by Bunčić (1991) 45% of pigs and 29% of cattle in Yugoslavia were found to harbour the pathogen. L. monocytogenes isolation rates from bovine and small-ruminant farm animal (goat and sheep) fecal samples in the US have been reported as 22% and 17%, respectively (Nightingale et al., 2004). In many cases, animals are merely carriers of L. monocytogenes and do not exhibit any signs of disease (Nightingale et al., 2004). Similarly, L. monocytogenes is also a transitory resident of the intestinal tract in humans, with 2–10% of the general population being carriers of the microorganism without any apparent  8  health consequences (Buchanan et al., 2017).  While L. monocytogenes incidence rates will differ across countries, farms, and animals, these examples demonstrate the ubiquitous nature of L. monocytogenes in our environment and likewise, our food system.   Consumption of contaminated foods is believed to be the principle cause of L. monocytogenes infection but direct transmission through contact with the environment, infected animals, or by cross-infection between patients is also a concern (Farber and Peterkin, 1991). Contamination of food can occur before harvest, especially fresh produce since soil or irrigation water can become contaminated by feces from animals such as birds, mammals, fish and invertebrates (Nightingale et al., 2004). After harvest, food materials can become contaminated in processing facilities, retail establishments, or consumer homes. However, literature most commonly reports contamination from food-processing facilities as being the primary cause of outbreaks (Tompkin, 2002). Here, L. monocytogenes has been found to survive in drains, standing water, residues, food-contact surfaces and floors for extended periods of time with the same genotype being re-isolated for years despite cleaning efforts and periods with inactivity (Tompkin, 2002; Wulff et al., 2006). Regardless, post-processing levels of L. monocytogenes in foods are often low and unlikely to cause disease. It is therefore products that support the growth of L. monocytogenes throughout their shelf-life that pose the largest concern to consumer health and these are typically RTE-foods such as soft cheeses, milk, vegetables, seafood, and processed meats. RTE-food categories implicated in L. monocytogenes outbreaks occurring in the US between 1998 and 2008 are shown in Table 1-1.      9  Table 1-1. Characteristics of 24 listeriosis outbreaks by implicated food categories, Foodborne Disease Outbreak Surveillance System, United States, 1998-2008. [Adapted from Cartwright et al. (2013) with permission] Food category No. outbreaks Total no. cases No. (%) hospitalized No. deaths (%) Deli meats 6 132 49 (37) 15 (11) Frankfurters 3 116 101 (87) 14 (12) Other meats 2 14 3 (21) 0 Mexican-style cheese 4 45 41 (91) 1 (2) Other dairy products 2 8 7 (88) 4 (50) Salad/other 3 27 7 (26) 3 (11) Unknown 4 17 7 (41) 1 (6) Overall 24 359 215 (60) 38 (11) 1.3.1.1 L. monocytogenes genomics L. monocytogenes has a singular, circular chromosome which is typically 2.7 to 3 Mbp in length (Cabanes et al., 2011; Glaser et al., 2001; Kuenne et al., 2013) with an average GC content (proportion of guanine and cytosine nucleotides within the genome) of 38% (Bohlin et al., 2010; Kuenne et al., 2013), placing them into the low GC content group of bacteria. The genomic GC content of bacteria varies from less than 20% to more than 70% and has been associated with properties such as genome size (Mitchell, 2007), oxygen, and nitrogen exposure (McEwan et al., 2004; Naya et al., 2002) and specific habitats (Chen and Zhang, 2003; Foerstner et al., 2005). For example, intracellular bacteria such as L. monocytogenes tend to have smaller genomes that are richer in AT compared to bacteria that live solely in soil such as Actinobacteria with their GC content of ~66% and 4.8 Mbp genomes (Wassenaar et al., 2009). The exact reasons for these associations remain putative (Hildebrand et al., 2010).  The first complete genome sequence of L. monocytogenes (EGDe) was released in 2001 (Glaser et al., 2001). Since then, a large number of L. monocytogenes strains have been sequenced and will continue to be sequenced for many years to come. The National Centre for Biotechnology Information (NCBI) database currently (Sept 22, 2017) contains a total of 137 complete L.  10  monocytogenes genomes and 1376 draft +scaffold assemblies that can be accessed through Genbank. Large contributors include academia and government agencies such as the US Centre for Disease Control (CDC) which launched a Listeria whole genome sequencing project in 2013 where all clinical and food isolates are sequenced to aid in outbreak investigations (US CDC, 2016). L. monocytogenes chromosomes contain around 2900 putative protein-coding genes, of which ~65% have an assigned function. A large majority of these genes encode putative transport systems, transcriptional regulators, and surface and secreted proteins, which allow the bacterium to adapt to diverse environmental conditions (Glaser et al., 2001). Transcriptome studies conducted on L. monocytogenes under various conditions commonly report that it expresses more than 98% of its ORFs at any given time and that these ORFs are organized into over 500 polycistronic operons which encompass ~60% of the annotated genome (Toledo-Arana et al., 2009).  Recent advances in molecular and bioinformatics techniques have discovered that in addition to coding sequences, bacterial and eukaryotic genomes alike also contain a number of non-coding RNAs (ncRNAs). ncRNAs exist in several different forms with most belonging to one of three main categories: 1) Cis-regulatory RNAs, 2) trans-encoded small RNAs (sRNAs), and 3) antisense RNAs (asRNAs) (Mellin and Cossart, 2012). Cis-regulatory RNAs are located at the 5’-ends of mRNA and fold into alternative structures in response to physicochemical cues. These transcripts are often referred to as riboswitches or thermosensors. Trans-encoded sRNAs, on the other hand, are not located adjacent to their target and share only limited complementarity allowing them to regulate multiple mRNAs. Lastly, asRNAs are transcribed from the DNA strand opposite of a gene and thus have perfect complementarity. L. monocytogenes has been shown to contain over 50 sRNAs, and over 100 asRNAs (Behrens et al., 2014; Christiansen et al., 2006; Mraheil et  11  al., 2011; Toledo-Arana et al., 2009; Wehner et al., 2014). Given the many types of ncRNAs, they exhibit a broad range of functions though studies have predominantly focused on their roles in transcription regulation and virulence.    L. monocytogenes genomes are also known to harbour a number of mobile genetic elements such as transposons, prophages, and mobile islands (Kuenne et al., 2013). For example, L. monocytogenes strains have been found to harbour plasmids with rates of plasmid detection ranging from as low as 0% in some studies to as high as 79% in others (Kolstad et al., 1992; Lebrun et al., 1992; McLauchlin et al., 1997; Perez-Diaz et al., 1982; Peterkin et al., 1992). The putative or known importance of these genetic elements will be discussed in sections to come. 1.3.1.2 L. monocytogenes subtypes  1.3.1.2.1 Genetic lineages and serotypes The original means by which Listeria spp. were, and continue, to be subtyped is based on the serological reactions of Listeria somatic (O-factor) and flagellar (H-factor) antigens with specific antisera (Paterson, 1940; Seeliger and Höhne, 1979).  This produces several serovars, also known as serotypes. L. monocytogenes contains a total of 13 serotypes (Seeliger and Jones, 1986) which are listed in Table 1-2. These serotypes can also be separated into four lineages (Table 1-2) using phylogenic and subtyping methods (den Bakker et al., 2008; Gray et al., 2004; Nightingale et al., 2005b; Orsi et al., 2008b; Ragon et al., 2008; Ward et al., 2008, 2004). Notably, strains belonging to the same serotype can belong to separate lineages. The majority of L. monocytogenes isolates cluster into lineages I (LI) and II (LII) and likewise over 90% of human listeriosis cases and outbreaks are associated with serotype 1/2a, 1/2b, and 4b strains (McLauchlin et al., 2004; Swaminathan and Gerner-Smidt, 2007). Figure 1-1 shows a summary of the distribution of L.  12  monocytogenes lineages among different ecological sources including that LI serotype 4b isolates are largely associated with human listeriosis outbreaks in the US whereas LII serotype 1/2a isolates are largely isolated from food and environmental sources (Orsi et al., 2011).  Table 1-2. Summary of L. monocytogenes lineages. Table adapted from Orsi et al. (2011) with permission. Lineage Serotypes Genetic characteristics Distribution I 1/2b, 3b, 3c, 4b, 4d, 4e Lowest diversity among the lineages, lowest levels of recombination. Isolated from various sources, overrepresented among human isolates. II 1/2a, 1/2c, 3a Most diverse, highest recombination rates.  Isolated from various sources, overrepresented among food-related and environmental isolates. III 4a, 4b, 4c, 7 Very diverse, recombination levels between those of lineage I and lineage II. Most isolates obtained from ruminants. IV 4a, 4b, 4c Few isolates analyzed to date.  Most isolates obtained from ruminants.  Figure 1-1. Distribution of L. monocytogenes lineages among different ecological compartments. The thickness of each arrow represents the proportion of isolates belonging to each lineage among isolates obtained from a given ecological compartment (e.g., outbreak, sporadic cases, animals, foods, and environment). Broken arrows indicate the proportion of isolates associated with cases in Northern Europe as here lineage II strains seem to be more common among sporadic cases than lineage I strains. Black solid arrows indicate the proportion of isolates associated with cases in the US. Lineages III and IV have been excluded due to their rare appearances. Figure and caption are from Orsi et al. (2011) and used with permission.  13  1.3.1.2.2 Clonal complexes and sequence types Beyond L. monocytogenes lineages and serotypes, the genetic relatedness of isolates can be further evaluated through sequence typing. The most commonly employed method of sequence typing is multilocus sequence typing (MLST) and involves sequencing the internal fragments of 6-14 housekeeping genes, which are essential for cell survival and reproduction (Maiden et al., 1998). Housekeeping genes generally evolve relatively slowly and thus are highly conserved. The L. monocytogenes MLST scheme is based on the sequencing of seven genes which include abcZ, bglA, cat, dapE, dat, ldh, and lhkA. MLST results have been shown to be congruent with pulsed-field gel electrophoresis (PFGE) results (Henri et al., 2016) which prior to whole genome sequencing, was considered the gold standard for strain typing (Graves and Swaminathan, 2001; Murchan et al., 2003).  While MLST is a great tool for studying the overall composition of bacterial populations, it lacks the ability to discriminate between closely related strains. To gain enhanced discriminatory power, Zhang et al. (2004) developed another MLST scheme that was solely based on virulence gene sequences (prfA, inlB, inlC, dal, lisR, clpP) and thus termed MVLST. This method has been shown to be more discriminatory for serotype 1/2a and 4b strains compared to MLST. However, MVLST has been rarely used in literature largely due to lack of international enforcement (Chen et al., 2007; Knabel et al., 2012; Sabat et al., 2013). The Institute Pasteur database (publicly available at: http://bigsdb.pasteur.fr/Listeria/Listeria.html) provides access to genotypic data for Listeria isolates and currently (Sept 22, 2017) contains 3318 isolates with 1262 different MLST profiles which are commonly referred to as sequence types (STs). Of the 1262 Listeria spp. STs, 1200 belong to L. monocytogenes. Although STs tend to be associated with specific L.  14  monocytogenes serotypes, it is also possible for isolates with the same ST to have different serotypes. MLST data can also subsequently be used to further classify isolates into clonal complexes (CCs). L. monocytogenes CCs are defined as groups of strains that share six out of seven MLST allelic profiles with at least one other member of the group (Feil, 2004). The Institute Pasteur database currently (Sept 22, 2017) contains 93 L. monocytogenes CCs. The term singleton is used to define STs having at least two allelic mismatches with all other STs.  Chenal-Francisque et al. (2011) were the first to survey the global clonal diversity of L. monocytogenes by analyzing 300 isolates from five continents. Here they found the worldwide dominance of the following clonal complexes arranged in descending order: CC2 > CC1 > CC3 > CC9 > CC7 > CC59 > CC121 > CC288 > CC199 > CC8 > CC101 > CC155 > CC5 > CC6. Similarly, some of the most commonly observed STs were ST2 > ST1 > ST9 > ST3 > ST121 > ST59 > ST7 > ST288 > ST5 > ST199 > ST145 > ST18 >ST6. Interestingly, a number of studies have reported that CC121 strains are frequently associated with foods and food-processing environments (Chenal-Francisque et al., 2011; Ebner et al., 2015; Martín et al., 2014; Maury et al., 2016; Parisi et al., 2010). This suggests that classifying strains based on their CCs may help us gain a better understanding of the diversity of L. monocytogenes in food processing environments.  1.3.1.3 L. monocytogenes pathogenicity As previously mentioned, the primary cause of listeriosis is the consumption of contaminated foods. Once L. monocytogenes has been consumed, its first challenge is to survive the harsh conditions of the stomach which includes pH levels between 2-5 and a number of  15  lipolytic and proteolytic enzymes. Cells that survive then continue along the gastrointestinal tract to the intestines (Fig 1-2). Here L. monocytogenes can cause acute febrile gastroenteritis in healthy individuals, with symptoms usually resolving spontaneously and without the need for medication. However, in people with weakened immune systems L. monocytogenes is able to enter into the host through the intestinal mucosa. In order for this to occur, specific ligands (internalins A and B) on the exterior cell wall of L. monocytogenes have to bind to receptors (Met and E-cadherin) on the exterior of enterocytes (Fig 1-3). Only specific versions of Met and E-cadherin will bind with the InlA and InlB proteins on cell wall of L. monocytogenes, and thus not all animals are affected by this pathogen.   Figure 1-2. Schematic representation of the physiopathology of L. monocytogenes infection [obtained from Vázquez-Boland et al. (2001) with permission].   16   Figure 1-3. Specificity of L. monocytogenes invasion proteins internalin A (InlA) and B (InlB) for human, guinea pig, and mouse and rat Met and E-cadherin intestinal cell receptors [obtained from Lecuit (2007) with permission].  Once across the intestinal barrier, the bacteria can spread rapidly via the lymph or blood, to the mesenteric lymph nodes, the spleen, and the liver (Fig 1-2). Not all L. monocytogenes cells are destroyed by the host macrophages, and surviving cells may proliferate in the liver (2-5 days). At this point the bacteria are usually eliminated by the immune system of healthy individuals; however, in immunocompromised individuals, unlimited proliferation of the bacteria in the liver parenchyma may result in the release of bacteria into the circulatory system (Vázquez-Boland et al., 2001). Once it is in the circulatory system the pathogen can cause septicemia. It can also then infect the central nervous system where it can cause encephalitis or meningitis. In pregnant women, L. monocytogenes can cross the placenta, leading to fetal infections and abortion (McLauchlin et al., 2004). Individuals with the invasive form of listeriosis usually report symptoms 1 to 4 weeks following the consumption of the contaminated food. The disease can be treated successfully with  17  antibiotics; however, the human mortality rate remains high (20-40%) despite these treatment options (US CDC, 2017). The minimal infectious dose for listeriosis is unclear and most likely varies between individuals; however, it is generally accepted that levels below 100 colony-forming units (CFUs) per gram of food are unlikely to cause disease (Health Canada, 2016). While all L. monocytogenes serotypes can potentially cause disease, 98% of human isolates belong to serotypes 4b, 1/2a, and 1/2b (Barbour et al., 2001). The reason these serotypes are more likely to cause disease is linked both to their prevalence in the environment and foods, and also to their genetic makeup.  Two clusters of genes are necessary for the invasion and intracellular replication of L. monocytogenes. The first cluster is referred to as ‘‘Listeria pathogenicity island 1’’ (LIPI-1), and encodes genes that allow the bacteria to infect cells and tissues, escape from the phagocytic vacuole, replicate within the cell, and then spread from cell-to-cell through actin-based motility (Chen et al., 2007). Notably, prfA, which encodes the major transcription regulator of the L. monocytogenes virulence regulon, is among the many genes within LIPI-1. The second cluster of invasion-related genes contains only two genes; inlA and inlB, which together make up an operon that is controlled by both PrfA and the RNA polymerase alternative sigma factor σB (McGann et al., 2007).  Several L. monocytogenes virulence-associated genes are either absent or exist in truncated forms in certain STs, CCs or on an individual strain basis (Nightingale et al., 2005a). For example, a number of additional virulence associated internalins aside from InlA and InlB, have been reported absent in a number of LIII strains, particularly internalins C, F, G, H, E, I, and J (Chen et al., 2011; Doumith et al., 2004; Hain et al., 2012; Jia et al., 2007; Nelson et al., 2004). Furthermore, premature stop codons (PMSC) have also been observed in prfA, plcA, and inlAB (Chen et al.,  18  2011; Nelson et al., 2004). LII strains in particular have a proportion of strains (>30%) that are virulence-attenuated due to PMSCs in inlA, whereas serotype 4b and 1/2b isolates typically harbour a full-length version of the gene (Poyart et al., 1996). Interestingly, the majority, if not all serotype 1/2c strains seem to carry inlA PMSCs (Jacquet et al., 2004; Tamburro et al., 2010). These PMSCs occur in the region of the translated protein that attaches to the exterior cell wall, thereby disabling the protein from attaching and likewise binding to E-cadherin on enterocytes, making these strains attenuated in virulence (Nightingale et al., 2005a). As a result, while serotype LII strains are more commonly found in foods, they are typically underrepresented among clinical isolates. Specific L. monocytogenes CCs have also been found to be associated with outbreaks. For example, Knabel et al. (2012) reported that the majority of Canadian human listeriosis outbreaks (1988–2010) have been caused by members of L. monocytogenes CC8 (ST 120), which belongs to serotype 1/2a. Similarly, L. monocytogenes strains belonging to ST7 (serotype 1/2a) and ST5 (serotype 1/2b) were associated with a large US multistate cantaloupe outbreak in 2011 and have also been reported to be prevalent in US chicken processing plants (Lomonaco et al., 2013). These findings suggest that certain genotypes may be better suited for survival in food processing environments or as previously discussed, contain genes or versions of genes that enhance their virulence potential.  1.3.2 L. monocytogenes physiology and food-related stress tolerances Overall, Listeria spp. are relatively nutritionally undemanding, requiring only biotin, riboflavin, thiamine, thioctic acid, and amino acids for optimal growth (Wagner and McLauchlin, 2008). Listeria also requires carbohydrates such as glucose for multiplication, from which acid is  19  produced as a by-product. Commonly recognized for its psychrophilic behaviour, L. monocytogenes can grow at temperatures as low as -0.4°C (Walker et al., 1990) and as high as 43°C (Doyle et al., 2001), but exhibits a maximum growth rate at around 30°C (Baranyi and Tamplin, 2004). Most strains are motile between 20 and 25°C though motility has also been observed at temperatures as low as 4°C (Cordero et al., 2016) and as high as 37°C (Gründling et al., 2004). Expression and repression of flagella synthesis is temperature and catabolite regulated and has been shown to be associated with virulence and both adherence and biofilm formation on abiotic materials (Lemon et al., 2007; Dons et al., 1992; Grundling et al., 2004; Kathariou et al., 1995).    Bacteria found in foods have three fates: growth, sustainability, or death. While growth poses the largest concern, it is equally important to understand how long pathogens can maintain viability in foods. Several factors affect the stress tolerance and survival of bacteria, including strain variation, age of the microorganisms, growth conditions (temperature, media), test conditions, and prior stress exposure (Doyle et al., 2001). In almost all cases, bacterial stress tolerance is growth-phase dependent with stationary phase cells demonstrating significantly enhanced stress tolerance compared to exponential-phase cells (Davis et al., 1996; Mackey et al., 1995; Phan-Thanh et al. 2000; Knabel et al., 1990; Welsh and Herbert, 1999; Cheroutre-Vialette et al. 1998). Thus, it is important to keep in mind that the growth and survival limits currently available in literature are only representative of the strains and conditions used in each individual study. The sections below will discuss the levels of food-related stresses that L. monocytogenes is capable of tolerating and some of the factors that can influence these levels.     20  1.3.2.1 Acid tolerance The optimal pH for L. monocytogenes growth is 7.1; however, it can maintain viability at pH levels between 3 and 12 (Liu et al., 2005) with between 4.6 and 10.5 typically permitting growth (Sorrells et al., 1989; Vasseur et al., 1999). Given these levels of acid tolerance it is not surprising that L. monocytogenes is able to survive passage through the stomach which has a typical pH range of 1.5-5.5 (Dressman et al., 1990).  The type of acid used to acidify a product is known to affect the growth abilities of L. monocytogenes. The inhibitory effect of acids can be correlated with their dissociation constant (pKa value), as cell membranes are known to be more permeable to weak acids due to in their undissociated form. Vasseur et al. (1999) found that acetic acid was the most effective against L. monocytogenes followed by lactic acid and then hydrochloric acid. Sorrells et al. (1989) also reported acetic acid to be the most effective followed by lactic acid, citric acid, malic acid, and hydrochloric acid. Therefore, it is important that food producers carefully select which acids to use in their products to ensure food safety. Temperature is also known to affect the acid tolerance of L. monocytogenes. Sorrells et al. (1989) found that acid stress survival (pH 5.0-5.6) was least at 35°C and greatest at 10°C. Likewise, Conner et al. (1986) reported that at pH 4.8 the death rate for L. monocytogenes was slower at 5°C than at 30°C, posing a greater concern for preserved, refrigerated, RTE-foods. Acid tolerance in L. monocytogenes is also greatly influenced by prior growth conditions. Kroll and Patchett (1991) found that prior growth at pH 5 substantially increased the survival of L. monocytogenes subjected to acid shock (20 min at pH 3) compared to cells pre-cultured at pH 7. Similarly, Koutsoumanis et al. (2003) evaluated the effect of food processing-related stresses on the acid tolerance of L. monocytogenes and found that exposure to mildly acidic media (pH 5  21  and 6) and growth in the presence of glucose resulted in enhanced survival at pH 3.5 but that prior exposure to sublethal levels of osmotic (10.5% NaCl), heat (45, 50°C), and low-temperature stresses (4°C) had no effect. In contrast, Faleiro et al. (2003) observed that osmoadaptation (3.5% NaCl, 2 h) induced cross-protection against acid shock conditions (pH 3.5). Cross-protection occurs when the metabolic and physiological changes that take place in a bacterium as a result of exposure to one stress, are beneficial for adapting to or surviving a subsequent stress. Such changes include the synthesis of stress response proteins and alterations in cell surface proteins and membrane lipid compositions (Begley et al., 2002; Leyer and Johnson, 1993; Lou and Yousef, 1997). Cross-protection is an important phenomenon to consider when predicting the survival of L. monocytogenes and other bacterial pathogens in both foods and during gastric passage. For example, L. monocytogenes from foods containing sublethal levels of salt or similar preservatives may exhibit cross-protection against low pH which may enable the bacterium to survive transit through the stomach into the small intestine where bacteria further encounter low oxygen, elevated osmolarity, and bile salts. Since it has been shown that acid adaptation can result in enhanced osmo- and bile salt tolerance (Begley et al., 2002; Ramalheira et al., 2010), passage through the stomach may further provide cross-protection against the harsh conditions present in the small intestine.  1.3.2.2 Cold and heat tolerance L. monocytogenes is capable of growing at temperatures between -0.4°C and 43°C and overall, is not widely considered to be heat tolerant. Golden et al. (1988) reported only minor losses in the viability of L. monocytogenes after 75 min at 50°C; however, after 10 min at 60°C no cells could be detected which was equivalent to a 6.5 log reduction. In Canada, milk is required to be pasteurized at 72°C for 15 s (Health Canada, 2005); however, Doyle et al. (1987) showed that  22  these conditions were ineffective in eliminating L. monocytogenes from milk contaminated with 2-4 log CFU/ml but found that 77°C for 15 s was sufficient. Despite this, L. monocytogenes outbreaks have not been associated with pasteurized milk products.  The composition of the media in which cells are grown and heated also has a large impact on heat stress survival. Casadei et al. (1998) showed that L. monocytogenes cells grown in a fat-rich medium (butter and cream) were four to eight times more resistant to heat at 60°C compared to cells grown in tryptic soy broth (TSB). Similarly, Jørgensen et al. (1995) found that L. monocytogenes cells grown in 1.5 M NaCl had significantly increased heat resistance (60°C) compared to cells grown in 0.9 M NaCl. Other food ingredients such as glucose, sucrose, and curing salts have also been reported to increase the thermal resistance of L. monocytogenes (Schoeni et al., 1991; Yen et al., 1991), highlighting the importance determining the thermal resistance of L. monocytogenes in individual food products and in products following any compositional changes.  Similar to the cross-protection examples shown for acid tolerance, pre-conditioning L. monocytogenes to other stresses prior to heat exposure can result in increased heat tolerance. Lou and Yousef (1996) found that pre-conditioning L. monocytogenes cells to pH 4.5 resulted in significantly enhanced heat tolerance at 56°C while Bunning et al. (1990) on the other hand found that pre-conditioning L. monocytogenes at sublethal temperatures of 35, 42, 48, and 52°C for various times had no impact on the bacterium’s survival at 58°C. Similarly, Farber and Brown (1990) reported that pre-conditioning cells at 48°C for 30 or 60 min prior to subjecting them to 64°C had no impact on survival; however, pre-conditioning for 120 min at 48°C resulted in a 2.4-fold increase in the time required for a 1 log reduction at 64°C. Starvation of cells (growth in minimal media) prior to heat exposure (56°C) has also been shown to significantly enhance the  23  thermotolerance of L. monocytogenes (Lou and Yousef, 1996). Such cross-protection situations commonly occur during food production and may have contributed to some foodborne outbreaks. For example, preservatives are often added to foods prior to thermal processing and some ingredients may be thermally processed before they added to a product that is then subsequently thermally processed as a whole. Additionally, many frozen meat products are pre-cooked and then re-heated by consumers prior to consumption.  L. monocytogenes is also known to be very cryotolerant and similar to heat stress, it’s cryotolerance is enhanced by the presence of food components.  El-Kest and Marth (1991) reported that 2% glycerol, lactose, milk fat, or casein enhanced the survival of L. monocytogenes during frozen storage at -18°C compared to phosphate buffer alone. However, regardless of a food’s composition, L. monocytogenes tolerates freezing temperatures extremely well. Golden et al. (1988) showed that during 14 days of storage in tryptic soy broth at -18°C, four L. monocytogenes strains exhibited no appreciable reduction in cell numbers.  While L. monocytogenes survives frozen storage, it grows when stored at refrigeration temperatures. To put the cold growth ability of L. monocytogenes into perspective, bologna containing minimal preservatives and initially contaminated with 10 CFU/g, can reach 4 log CFU/g by the end of 4-weeks of storage at 4°C, and reach 8 log CFU/g by the end of 8-weeks (Seman et al., 2002). While the addition of preservatives in deli meats usually prevents L. monocytogenes from reaching unsafe levels (Seman et al., 2002), the increasing consumer demand for fresh, unprocessed RTE-foods makes it challenging for the food industry to ensure food safety. Additionally, household refrigerators commonly fluctuate in temperature leading to temperature abuse and permitting L. monocytogenes to reach higher levels. For example, Burnett et al. (2005) showed that after 5-weeks of storage at 5°C, L. monocytogenes levels on roasted turkey slices  24  increased by 2 log CFU/g whereas after 5-weeks at 10°C it had increased by 5-log CFU/g. Furthermore, various compounds have been shown to enhance the cold growth abilities of L. monocytogenes such as the addition of the compatible solutes glycine betaine and carnitine (Ko et al., 1994; Smith, 1996) which can be found in meat products. Similarly, the inclusion of short peptides has also been shown to confer a similar effect (Verheul et al., 1998). The mechanisms behind these cryoprotectants are discussed in section 1.3.4.2.1. 1.3.2.3 Osmotolerance Osmotic stress is defined as an increase or decrease in the osmotic strength of the environment surrounding an organism (Csonka, 1989). While elevated salt solutions are most commonly associated with osmotic stress, other solutes such as sugars can also impose hyperosmotic stress on organisms whereby water, through the process of osmosis, diffuses out through the cell membrane leaving cells dehydrated. As a consequence of the decreased cytoplasmic volume, the concentrations of intracellular metabolites increase and thus cause a reduction in intracellular water activity (Csonka, 1989). To counteract the flow of water out of cells, organisms increase their internal concentration of specific solutes through either de novo synthesis or membrane transport. For example, the external presence of glycine betaine and carnitine which are taken up by the cell via membrane transport proteins (Smith, 1996; Ko et al., 1994) have been shown to enhance the osmotolerance of L. monocytogenes. In addition to counteracting the flow of water out of the cell, these molecules can help maintain or enhance protein stability under stress conditions (Ali et al., 2017; Wesche et al., 2009). Since the molecules accumulated do not impact or inhibit cellular processes, they are referred to as compatible solutes (Csonka, 1989).    25  L. monocytogenes is well known for being able to tolerate high levels of salt and grow in salt solutions up to 14% (w/v) (Farber et al., 1992; Shabala et al., 2008). Thus, food preservation methods usually include salt in combination with mild acid levels, and refrigeration to ensure that this pathogen cannot reach harmful levels in food. Farber et al. (1992) determined the minimal water activity requirements for the growth of L. monocytogenes using NaCl, sucrose, and glycerol. They found that glycerol was the least toxic of the three solutes with three of five strains capable of growing in BHIB adjusted with glycerol to an aw of 0.90 at 30°C compared to a minimum aw of 0.93 and 0.92 in broth adjusted with sucrose and NaCl, respectively. Notably, L. monocytogenes is one of the few foodborne pathogens that can grow at an aw value below 0.93. In the same study, they also found that the minimum aw value required for growth generally increased as the incubation temperature decreased, demonstrating that temperature also plays an important role in the osmotolerance of L. monocytogenes. Similar results were found by Cole et al. (1990) who reported that at 5°C L. monocytogenes could only grow in up to 8% NaCl compared to 12% at 30°C. Faleiro et al. (2003) also observed that a lower temperature (8 vs. 30°C) was associated with a decrease in resistance to 2% NaCl. In another study by Shahamat et al. (1980), L. monocytogenes was able to survive for 4 days in TSB containing 25.5% NaCl at 37 °C, 24 days at 22 °C, and more than 132 days at 4 °C. In this study, survival in a hyperosmotic environment is extended at lower temperatures.  Again, cross-protection can also influence the osmotolerance of L. monocytogenes. In a study by Lou and Yousef (1997), adaption to ethanol (5% v/v) and heat shock (45°C, 1 h) significantly increased the resistance of L. monocytogenes to 25% NaCl whereas adaptation to acid stress (pH 4.5-5.0, 1 h) did not confer significant protection against osmotic stress conditions. Contrary to this, Faleiro et al. (2003) found that adaptation to acid stress (pH 5.5, 2 h) induced a  26  cross-protection against osmotic shock conditions [20% (w/v) NaCl]. Similar findings were also reported by O’Driscoll et al. (1996). Cross-protection poses a concern for food manufacturers such as the dairy industry, where both salt and acid are used. For example, in cheese production an increase in acidity occurs first, which may induce an acid tolerance response in the pathogen that then allows it survive for a longer period of time once it enters the cheese brine solution.  1.3.2.4 Desiccation tolerance Desiccation tolerance reflects a bacteria’s ability to survive on a surface for extended periods of time with little access to nutrients and water. As so, desiccation tolerance is believed to be associated with the ability of L. monocytogenes to persist in food production plants (Vogel et al., 2010) and subsequently cross-contaminate food products. L. monocytogenes has been shown to survive desiccation for three months in a simulated food-processing environment (Vogel et al., 2010). Furthermore, the same strains of L. monocytogenes have been repeatedly isolated from the same processing plants for up to eight years, despite proper sanitation efforts (Gudmundsdóttir et al., 2006; Keto-Timonen et al., 2007; Møretrø and Langsrud, 2004). L. monocytogenes can also encounter desiccation stress in or on foods such as on the exterior of an apple, or in dried products such as grains, fruit leathers, and jerkies.   Relative to other stresses, desiccation tolerance has been largely understudied in L. monocytogenes; however, a few publications have reported the influence of temperature, humidity levels, organic residues, salt, and biofilm formation. Palumbo and Williams (1990) found that during drying, the survival of L. monocytogenes was higher at 5°C than at 25°C whereas the impact of different RHs (1-75%) on survival was less pronounced. This suggests that L. monocytogenes  27  is likely able to survive longer on dry surfaces in refrigerated food manufacturing plants compared to those kept at ambient temperature.  The presence of organic matter also enhances the desiccation survival of L. monocytogenes, thereby increasing the risk of cross-contamination within a food processing facility. L. monocytogenes dried on glass coverslips showed enhanced survival when suspended in beef extract, glycerol, corn syrup, skim milk, and canned milk compared to distilled water (Palumbo and Williams, 1990). Similarly, Helke et al. (1994) found that the desiccation survival of L. monocytogenes and Salmonella Typhimurium at 25°C and 32.5 or 75.5% RH was increased in dilute pasteurized whole milk compared to PBS. Takahashi et al. (2011) air dried L. monocytogenes, S. aureus, and S. Typhimurium at 25°C on stainless steel coupons previously coated in minced tuna, ground pork, and cabbage. They observed that minced tuna and ground pork increased the survival of all bacteria types during the first 14 days of the 30-day experimental period with L. monocytogenes showing the greatest survival. Smoked and non-smoked salmon juice has also been shown to increase the desiccation (43% RH) survival of L. monocytogenes strains on stainless steel coupons (Vogel et al., 2010). Similarly, both L. monocytogenes (Vogel et al., 2010) and Salmonella spp. (De Cesare et al., 2003) exhibited increased survival when desiccated in TSB as compared to physiological peptone saline or PBS. Given the similar impacts of both desiccation and osmotic stress on intracellular water levels, several studies have focused on the role of prior osmoadaptation and the presence of salt during desiccation, on the survival of L. monocytogenes. Such situations can occur when a preserved food mixture containing L. monocytogenes, is unknowingly spilled onto a food-processing surface and left unaddressed for a long period of time. Dreux et al. (2008) reported that L. monocytogenes had increased desiccation survival at 48% RH on parsley leaves when cells were  28  pre-cultured in 3% NaCl. Vogel et al. (2010) also reported enhanced desiccation survival when L. monocytogenes cells were pre-cultured in 5% NaCl and even higher survival counts when cells were subsequently desiccated in 5% NaCl.  The desiccation survival of L. monocytogenes is also greatly improved when cells are surrounded in a biofilm matrix (Hingston et al., 2013; Truelstrup Hansen and Vogel, 2011). This can occur when surface attached biofilm forming cells are initially exposed to a high moisture environment where upon biofilm development takes place and is then followed by an extended period under low RH conditions. Biofilm polymers are very hydrophilic and may aid in reducing the rate of biofilm and cellular drying. Studies have shown the bacterial transfer from biofilms to foods (cheese, bologna, salami) was enhanced when biofilms were first dried as opposed to wet (Rodriguez et al., 2007). This demonstrates that not only do biofilms enhance bacterial survival during desiccation, but dry biofilms are more successful at cross contaminating foods.  While this review will not cover L. monocytogenes biofilms in great detail, it remains an important factor contributing to the survival and cross-contamination of L. monocytogenes in food production facilities. Biofilms are commonly defined as communities of microorganisms attached to a surface or interface that produce an extracellular matrix of polymeric substances (EPS) in which cells are embedded (Costerton et al., 1995). They can consist of single or multiple microbial species with the latter being predominant in most environments. The components of EPS differ between bacterial species but it is generally accepted that the main component is extracellular polysaccharides but may also include proteins, lipids, nucleic acids, and other biopolymers (Flemming and Wingender, 2010; Harmsen et al., 2010). L. monocytogenes is capable of forming biofilms in areas of food-processing plants that are often difficult to clean such as such as air or liquid filtration systems, rubber or Teflon seals, and machine joints (Kumar et al., 1998; Lee Wong,  29  1998). In addition to protecting cells from desiccation, biofilms also enhance the resistance of cells to UV light, and treatment with antimicrobial and sanitizing agents (Pozos et al., 2004; Pan et al., 2006). 1.3.3 Factors associated with L. monocytogenes phenotypes important to the food industry 1.3.3.1 Strain origin One of the ways in which the stress tolerance of L. monocytogenes isolates is compared is based on their source or origin. While this means of differentiation is still used today and allows for a more applied understanding of experimental outcomes, it was more relevant prior to the availability of serotyping and more advanced molecular subtyping methods. Most of the differences observed between clinical and animal or food isolates can be linked to their serotype and as previously mentioned, clinical isolates more commonly belong to the LI serotypes 4b and 1/2b whereas environmental and food isolates more commonly belong to the LII serotypes 1/2a, 1/2c, and 3a. However, as previously alluded to, most clinical isolates merely represent a subset of more virulent food isolates and therefore the outcomes from these studies are often inconclusive. Nonetheless, some consistent trends have been observed. In a study by Dykes and Moorhead (2000), two out of 15 L. monocytogenes strains isolated from meat were acid sensitive and none were acid tolerant, whereas all 15 clinical isolates were acid tolerant. Similarly, Ramalheira et al. (2010) characterized the ability of 27 strains of L. monocytogenes from food (n=16) and human listeriosis patients (n=11) to survive simulated gastrointestinal tract conditions and found that clinical strains had higher levels of survival relative to food strains. In line with these observations, serotype 4b strains have been shown to be more acid tolerant compared to serotype 1/2a strains. It is believed that the increased acid tolerance of  30  4b strains is likely a contributing factor to the prevalence of this serotype among clinical isolates, in addition to the higher incidence of full length inlA within this serotype. However, other studies have reported no difference in the survival characteristics of clinical and food isolates. In a study by Uyttendaele et al. (2004), no difference was found between the ability of seven clinical strains and seven food isolates to grow in several single and combined stresses involving different acidities, salt levels, and cold stress temperatures.   In a study by Vialette et al. (2003) the acid and osmotolerance of a number of L. monocytogenes strains responsible for human cases of listeriosis associated with contaminated seafood, was compared to L. monocytogenes strains isolated from similar seafood products. The study employed a factorial design involving the interactions of pH (5-7), NaCl concentration (0.5-8% v/v) and temperature (10 and 20°C). No differences were observed between clinical and seafood strains when testing acid or osmotolerance alone. However, when a pH≤6.0 was combined with 4% or 8% NaCl, the generation times of the clinical strains were faster than those of food strains. This supports the finding that clinical isolates are better suited to survive simulated gastric conditions compared to food isolates (Ramalheira et al., 2010) as L. monocytogenes likely experiences both salt and acid stress simultaneously during gastrointestinal travel.  Durack et al. (2013) evaluated the salt (12.5% NaCl) and cold (4°C) tolerance of 26 clinical isolates, 30 food-related isolates, and 85 isolates of animal origin. Here they reported that the majority of isolates capable of growing in 12.5% NaCl were from animal sources whereas clinical isolates exhibited significantly slower growth rates at 4°C compared to animal or food isolates. Again, this correlates well with serotype studies which have shown that serotype 1/2a strains have faster growth rates at low temperatures compared to 4b strains (Barbosa et al., 1994; De Jesús and Whiting, 2003).   31  Avery and Bunčić (1997) investigated the impact of a temperature change from 4°C to 37°C on the lag phase duration of L. monocytogenes strains from clinical (n=15) and meat (n=15) sources to simulate the ingestion of L. monocytogenes via refrigerated RTE food products. The results showed that clinical strains had significantly shorter lag phases than the meat strains following the temperature up-shift. Taken together with the findings of Durack et al. (2013), this suggests that while clinical strains have slower growth rates at 4°C, they are able to recover faster within the host organism.  Lastly, Begot et al. (1997) compared the growth abilities of 58 L. monocytogenes strains from meat products (n=43) and from meat and dairy plant environments (n=15) using a factorial design that included temperature (10 and 37°C), pH (5.6 and 7.0), and aw (0.96 and 1). The results showed that the majority of strains isolated from meat and dairy plant environments had faster growth rates than the meat product isolates under all conditions tested suggesting that environmental isolates may be better suited for surviving more diverse conditions compared to isolates found in foods.  1.3.3.2 L. monocytogenes subtypes Phenotype comparisons are also frequently made among L. monocytogenes subtypes with the majority of studies focusing primarily on strain lineages and serotypes. As mentioned previously, serotype 1/2a strains are commonly found to be more cold tolerant than 4b strains (Bunčić et al., 2001; Junttila et al., 1988; Lianou et al., 2006) and serotype 4b strains are generally more acid tolerant than 1/2a strains (van der Veen et al., 2008). However, it is important to note that in many cases these are merely generalized observations as opposed to significant differences, as many studies utilizes small strain sets. Barbosa et al. (1994) reported that out of 39 L.  32  monocytogenes strains, Scott A, a 4b clinical isolate, grew the slowest at 4°C and that 1/2a strains grew the fastest followed by 1/2b, and 4b strains. Similarly, De Jesús and Whiting (2003) observed that LII isolates (all serotype 1/2a) exhibited the shortest lag phase durations at 5°C followed by LI and then LIII isolates. More recently, researchers have suggested that LII strains may be able to survive better under food-related stresses due to an enhanced ability to acquire advantageous mutations and extrachromosomal DNA compared to LI strains which typically have more conserved genomes (Dunn et al., 2009; Orsi et al., 2007, 2008a, 2011; Ragon et al., 2008). More details regarding these genetic differences will be discussed in sections 1.3.3.6 and 1.3.3.7.    Studies have also elucidated differences between the salt tolerance of L. monocytogenes subtypes. Specifically, LI strains have been shown to be more salt tolerant than LII strains (Bergholz et al., 2010) and similarly, serotype 4b strains to be more salt tolerant than serotype 1/2a and 1/2b strains (Bergholz et al., 2010; Ribeiro et al., 2014; van der Veen et al., 2008). Bergholz et al. (2010) analyzed the salt tolerance of 40 strains from the three major L. monocytogenes lineages and found that at 37°C, in BHIB with 6% NaCl, lineage I and III strains grew significantly faster than lineage II strains. Again, the association of enhanced salt tolerance with serotypes commonly associated with listeriosis may suggest that these strains are better able to survive gastric passage. Similarly, LI strains may also have faster growth rates in preserved food items; however, most preserved items are also refrigerated and LI isolates are generally considered to be cold sensitive relative to LII strains. Lianou et al. (2006) analyzed the heat and acid tolerance of 25 L. monocytogenes strains from various serotypes and found that serotype 4b strains had lower heat tolerance (55°C, 240 min) than did isolates representing all other serotypes combined, and that outbreak-related 4b strains had significantly lower acid (pH 3) death rates (higher acid resistance) than did the rest of the strains belonging to this serotype.  33  Perhaps the most studied phenotypes in relation to L. monocytogenes subtypes is the bacterium’s biofilm forming abilities. Borucki et al. (2003) reported that LII strains (1/2a and 1/2c), formed denser biofilms than LI strains. Specifically, serotypes 3a, 1/2c, and 1/2a had the highest average absorbance values following crystal violet staining, but the range of values was high and no statistically significant differences could be concluded. Another study evaluated the biofilm forming abilities of 92 L. monocytogenes strains, and found that 1/2a strains produced significantly more biofilm than all other serotypes (1/2b, 1/2c, 3a, 4a, 4b, 4e) (Rolf et al., 2011). In contrast, a similar study by Djordjevic et al. (2002) found that LI (4b and 1/2b) formed significantly more biofilm than LII (1/2a, 1/2c) and LIII (4a, 4b, 4c) strains. On the other hand, Kalmokoff et al. (2001) reported that among 36 L. monocytogenes strains, serotype (1/2a, 1/2b, 1/2c, 4b, and 4c) was not associated with biofilm forming abilities. These discrepancies are known to be a result of the use of different biofilm assessment assays as researchers are still working on finding the most accurate way to evaluate the biofilm forming abilities of L. monocytogenes.    1.3.3.3 Persistent vs. sporadic strains Some studies have been interested in determining whether persistent strains of L. monocytogenes which are consistently isolated from the same environment for a long period of time (months to years), exhibit different stress survival behaviour compared to strains that are sporadically observed in these environments. Lundén et al. (2000) evaluated the adherence of three persistent and 14 non-persistent L. monocytogenes strains (from poultry and ice cream plants) to stainless steel surfaces and found that persistent strains showed 2- to 11-fold higher adherence than non-persistent strains after 1 and 2 h contact times, respectively. Similarly, Borucki et al. (2003) found that persistent strains (n=11) from milk tanks exhibited increased biofilm production relative  34  to non-persistent (n=15) strains. These studies suggest that L. monocytogenes attachment and subsequent biofilm forming abilities are related to strain persistence in the food industry. Another study by Lundén et al. (2003) looked at the resistance and cross-adaptive responses of two persistent and two non-persistent strains of L. monocytogenes to two quaternary ammonium compounds. While one of the persistent strains showed increased resistance to both quaternary ammonium compounds, the limited number of strains evaluated in this study makes it difficult to extrapolate the findings. Fox et al. (2011) on the other hand, observed that persistent strains (n=4) had a higher level of resistance to two quaternary ammonium compounds compared to non-persistent strains (n=9). In contrast, a similar study conducted by Kastbjerg and Gram (2009) found no difference between the disinfection susceptibilities of persistent (n=5) and non-persistent (n=9) strains. Differences have also been observed in the food-related stress tolerance of persistent and non-persistent strains. Lundén et al. (2008) investigated the acid (pH 2.4, 2 h) and heat (55°C for 40 min) tolerance of 17 persistent and 23 non-persistent L. monocytogenes strains recovered from three meat-processing plants and found that persistent strains were significantly more acid tolerant whereas no differences were observed with regards to heat tolerance. Porsby et al. (2008) evaluated the ability of a persistent L. monocytogenes strain from fish smokehouses to survive a series of hurdles associated with the production of cold smoked salmon (salting, drying, and cold-smoking steps). They reported that the persistent strain did not displayed enhanced tolerance to the processing steps compared to a clinical strain or the reference strain EGD.  In conclusion, while some examples of phenotypes associated with persistent strains have been presented, overall most existing literature does not support the presence of an enhanced stress tolerance among persistent L. monocytogenes subtypes.  35  1.3.3.4 Presence or absence of genes The phenotypic characteristics previously discussed to be associated with L. monocytogenes lineages and serotypes are often attributed to the presence, absence, or genetic variation of genes. Overall, strains from lineages II, III and IV show higher rates of recombination and lower degrees of sequence similarity compared to lineage I strains which typically have more conserved genomes (Dunn et al., 2009; Orsi et al., 2007, 2008a, 2011, 2008a; Ragon et al., 2008). This is proposed to be due to LI strains living in less diverse environments. It has also been suggested that LI represents descendants of a recently emerged highly virulent clone (den Bakker et al., 2008; Orsi et al., 2008). Certain stress response genes, predominantly those involved in membrane transport and cell wall structure (Doumith et al., 2004), have been reported as present in LII isolates but absent among LI isolates (Borucki and Call, 2003; Call et al., 2003; Chan and Wiedmann, 2008; Doumith et al., 2004; Zhang et al., 2003). Given the critical roles of these structures in allowing bacteria to adapt and tolerate numerous stresses (Álvarez-Ordóñez et al., 2008; Annous et al., 1997; Klein et al., 1999; Verheul et al., 1997; Weber et al., 2001), it is not surprising that L. monocytogenes lineages and serotypes can behave differently under certain stresses. In a whole genome comparison study conducted by Zhang et al. (2003), LII strains were found to contain 16 lineage-specific regions that include 47 different genes that were absent in all LI strains. These genes fell into six different functional categories including cell wall components, small molecule transport, central metabolism, transcription regulation, and stress resistance. The largest functional category represented was cell wall components and included genes involved in teichoic acid biosynthesis, and genes encoding cell-wall anchored proteins. In the same study, three genes specific to serotype 1/2a strains, encoded proteins with leucine-rich repeats characteristic of the internalin family of  36  proteins. This leucine-rich repeats region allows internalins to bind structurally unrelated ligands, thereby implicating them in a wide range of functions (Kobe and Kajava, 2001) which may contribute to the diverse survival characteristics of serotype 1/2a strains. Other notable serotype 1/2a specific genes included a histidine kinase and its response regulator (lmo1060, lmo1061) that together make up a two-component regulatory system which allows organisms to quickly sense and respond to changes in their environment (West and Stock, 2001), and proAB which are responsible for L-proline biosynthesis and have been shown to offer L. monocytogenes osmotic protection in the absence of the preferred osmoprotectants betaine and carnitine (Sleator et al., 2001).  The alternative sigma factor, σC, and lmo1078, encoding a UDP-glucose pyrophosphorylase, are examples of genes with reported roles in L. monocytogenes cold tolerance, that are present in LII strains but absent in LI and serotype 4b strains, respectively (Chan and Wiedmann, 2008; Chassaing and Auvray, 2007). These absences may partly explain the reduced cold tolerance of serotype 4b isolates respective to 1/2a isolates. Similarly, van der Veen et al. (2008) hypothesized the enhanced acid tolerance of 4b strains relative to 1/2a strains may be due to the presence of ORF2110 which was elucidated by Doumith et al. (2004) to be specific to serotypes 4b, 4d, and 4e. ORF2110 resembles the serine protease HtrA that has been shown to be important for virulence and for growth at low pH, high temperature, and high osmolarity (Stack et al., 2005; Wilson et al., 2006; Wonderling et al., 2004). Chen et al. (2009) were also able to conclude that the gene lmo0038 which is absent from lineage III strains, has roles in the acid and heat stress responses of L. monocytogenes. Specific gene clusters have also been shown to influence L. monocytogenes phenotypes. One example is a five-gene cluster (lmo0444 – lmo0448) known as stress survival islet 1 (SSI-1)  37  which has been shown to contribute to the acid, salt, and low temperature stress tolerance of L. monocytogenes (Cotter et al., 2005; Ryan et al., 2010). Specifically, gadD1 and gadT1 (lmo0447 and lmo0448) which encode a glutamate decarboxylase system, as well as pva (lmo0446), which encodes a bile salt hydrolase (Begley et al., 2005), have been shown to be responsible for the enhanced tolerances to the aforementioned stresses. Hein et al. (2011) investigated the prevalence of SSI-1 among 117 L. monocytogenes strains from various serotypes and found it to be present in the majority of 1/2c, 3b, and 3c strains tested but rarely associated with serotypes 1/2a, 4a, 4b, 4d, and 4e. Recently, Harter et al. (2017) reported the presence of an additional stress survival islet (SSI-2) predominantly in L. monocytogenes ST121 strains (91%, n=51) which are commonly found in foods and food-processing environments. The islet contains two genes with homology to a transcriptional regulator (lin0464) and a PfpI protease (lin0465) which were shown to be activated under alkaline and oxidative stress conditions with lin0464 regulating the transcription of lin0465. Examples such as these are contributing factors to the differences observed in the phenotypic behaviour among strains from the same serotypes.   1.3.3.5 Differential expression of genes In addition to the presence or absence of genes affecting strain phenotypes, the same gene can also be regulated differently across strains. This may stem from promoter region differences which subsequently can impact the binding and transcription efficiency of RNA polymerases, or it may be the result of other factors such the presence of small RNAs which can block or enhance transcription.  Severino et al. (2007) compared the whole-genome expression profiles of six L. monocytogenes isolates of serotypes 4b, 1/2b, and 1/2a in late exponential phase cells grown in  38  defined media at 37°C and found that the virulence associated genes plcA, plcB, gly, actA, and inlAB were highly expressed in LI strains compared to their levels of expression in LII strains. A number of cell wall-associated genes found in both serotype 1/2 and 4 strains were also found to be overexpressed in LI strains compared to LII strains. Furthermore, sigB was overexpressed in LI, along with its operon genes rsbV, rsbW, and rsbX. Accordingly, 20 genes belonging to the σB regulon were upregulated in LI but not LII strains. LII strains on the other hand, were found to exhibit overexpression of motility- and chemotaxis-related genes. Such differences in gene expression may account for the higher prevalence of 4b strains among clinical isolates.  Gene expression differences have also been observed among strains with different levels of cold tolerance. Cordero et al. (2016) compared the global transcriptomes of two cold sensitive and two cold tolerant L. monocytogenes strains at 8°C and found that the most defining difference between them was the repression of motility-related genes in the cold tolerant strains, whereas these genes were upregulated in the cold sensitive strains. The authors hypothesized that the suppression of motility-related genes may allow cold tolerant strains to achieve a greater rate of proliferation at low temperature. The cold sensitive strains on the other hand, exhibited the upregulation of a greater number of genes involved in translation, ribosomal structure, and biogenesis compared to the cold tolerant strains. Similarly, Arguedas-Villa et al. (2010) observed that three cold tolerant strains had increased expression of two known cold-stress response genes (cspA and pgpH) compared to three cold sensitive strains.   Bowman et al. (2010) compared the gene expression responses of an acid sensitive and acid tolerant L. monocytogenes strain (both serotype 1/2a) exposed to pH 5 (HCl) and pH 5 in combination with 21 mM sodium diacetate. Their results showed that the σB activated operon pdhABCD encoding the pyruvate dehydrogenase complex (PDC), exhibited large increases in  39  expression in the acid tolerant strain and smaller increases in expression in the acid sensitive strain. The PDC converts pyruvate to acetyl coenzyme A (acetyl-CoA) and is a key metabolic enzyme important for aciduric growth (Chaturongakul et al., 2008). In the same study, they also compared the transcriptomes of these strains at pH 5 and pH 7 and found that 31% of σB-activated genes were upregulated in the acid tolerant strain whereas only 22% were upregulated in the acid sensitive strain. 1.3.3.6 Horizontal gene transfer  Despite the identification of individual or small clusters of genes associated with specific L. monocytogenes lineages or serotypes, the majority of genomic diversity among L. monocytogenes genomes is a result of horizontally transferred elements such as plasmids, phage insertions, and transposable elements. The acquisition of foreign DNA is fundamental to the evolution and diversification of most bacterial species (Brüssow et al. 2004; Frost et al., 2005; Hacker and Carniel, 2001; Recchia and Hall, 1997). Horizontal gene transfer is particularly important for many human pathogens as these integrative elements can encode functions that increase bacterial fitness under certain environmental conditions or may enhance microbial virulence (Benedek and Schubert, 2007; Brussow et al., 2004; Hacker and Carniel, 2001). Accordingly, it is important to monitor both the prevalence and spread of horizontally transferred elements among foodborne pathogens such as L. monocytogenes.  In a study by Hain et al. (2006), around 25% of the genetic regions absent in L. welshimeri relative to L. monocytogenes were due to L. monocytogenes acquiring horizontally transferred genes as opposed to L. welshimeri being subject to gene loss. In another study by Kuenne et al. (2013), the authors estimated that a third of the accessory genes in L. monocytogenes have been  40  introduced by mobile genetic elements. Specifically, the majority of differences among the accessory genomes of 11 L. monocytogenes strains from six serotypes were due to nine hyper variable hotspots, several prophages, three transposons (Tn916, Tn554, IS3-like), and two mobilizable islands.  Horizontally transferred elements are commonly identified by the presence of mobile element proteins such as integrases, recombinases, transposases, and resolvases. Additionally, horizontally transferred DNA may have a different GC content than the host genome as a result of different evolutionary pressures (Coburn et al., 2007; Hacker and Carniel, 2001). However, over time the GC content of these elements increasingly become more similar to the host genome making such regions more difficult to detect over time (Hain et al., 2006). 1.3.3.6.1 Plasmids L. monocytogenes strains are known to harbour plasmids with rates of isolation reaching as high as 79% (Dykes et al., 1994; Fistrovici and Collins-Thompson, 1990; Kolstad et al., 1992; Lebrun et al., 1992; Perez-Diaz et al., 1982; Peterkin et al., 1992). While the majority of strains only harbour one plasmid, some Listeria spp. have been found to contain two (Earnshaw and Lawrence, 1998; Margolles et al., 1998). Plasmid sizes typically range from to 30 to 100 kbp (Kuenne et al., 2010; Liang et al., 2016; Poyart-Salmeron et al., 1990) with GC contents around 36% (Liang et al., 2016; Schmitz-Esser et al., 2015) which is slightly lower than the average GC content of L. monocytogenes genomes (37-38% GC) (Kuenne et al., 2013).  To date, plasmids acquired by L. monocytogenes have been shown to contain genes that confer resistance to sanitizers, (Elhanafi et al., 2010; Katharios-Lanwermeyer et al., 2012; Rakic-Martinez et al., 2011), heavy metals (Katharios-Lanwermeyer et al., 2012; Lebrun et al., 1992;  41  Rakic-Martinez et al., 2011), and to a lesser extent, antibiotics (chloramphenicol, clindamycin, erythromycin, streptomycin and tetracycline) (Hadorn et al., 1993; Poyart-Salmeron et al., 1990). Listeria spp. plasmids also commonly contain several other uncharacterized efflux pumps including multiple drug resistance (MDR), small multidrug resistance (SMR) and multi antimicrobial extrusion (MATE) proteins (Boylan et al., 2006; Gibson et al., 2000; Kuroda and Tsuchiya, 2009; Masaoka et al., 2000). Additionally, a number of oxidative stress response genes (peroxidases, reductases) (Kuenne et al., 2010; Liang et al., 2016) have also been observed on L. monocytogenes plasmids but their exact roles have yet to be investigated. In other bacterial species, multidrug efflux pumps have been linked to stress response, virulence, and quorum sensing [which were reviewed by Li and Nikaido (2009)]. Ma et al. (1995) reported that transcription of a MDR pump (acrAB) in E. coli increased in response to fatty acids, ethanol, high salt, and cellular transitioning into stationary phase. While it has not yet been shown, it is likely that genes located on L. monocytogenes plasmids have active roles in facilitating the pathogens’ growth or survival in foods. In agreement with this, higher rates of plasmid harbourage have been reported among food and environmental isolates compared to clinical isolates (Lebrun et al., 1992; McLauchlin et al., 1997). Likewise, LII isolates have been shown to have higher rates of plasmid harbourage compared to LI isolates (Kolstad et al., 1992; Lebrun et al., 1992; Margolles et al., 1998; McLauchlin et al., 1997; Orsi et al., 2011). Furthermore, plasmids have been found to be highly conserved among specific L. monocytogenes sequence types commonly observed in food-processing environments, suggesting that a strong selective pressure is acting on them and that plasmids may facilitate adaptation to food-related stresses (Schmitz-Esser et al., 2015).     42  1.3.3.6.2 Genomic islands Genomic islands are defined as large genomic regions typically 10-200kb in length with probable horizontal origins (Langille et al., 2010). This broad definition then also includes other mobile genetic elements such as prophages, integrons, conjugative transposons and integrative conjugative elements (Langille et al., 2010). Often, genomic islands are associated with a specific function such as 1) the ability to utilize novel carbon and nitrogen sources (metabolic islands), 2) the ability to break down novel compounds (degradation islands), 3) resistance to antibiotic and heavy metals (resistance islands); and 4) the ability to cause disease (pathogenicity islands) (Dobrindt et al., 2004; Hacker and Carniel, 2001). Traditional genomic islands are characterized by the presence of an integrase, specific attachment sites (attL and attR) (Boyd et al., 2008), and insertion at a tRNA locus. However, this is not always the case and some genomic islands are inserted elsewhere in bacterial genomes and are devoid of any mobile elements.  A few well-described genomic islands have been identified in L. monocytogenes, namely Listeria pathogenicity islands 1, 2 and 3 (LIPI-1,2,3), and Listeria genomic islands 1 and 2 (LGI1,2). LIPI-1 consists of three transcriptional units which include the virulence essential genes hly, mpl, actA, plcB, prfA and plcA [reviewed in Vazquez et al. (2001)]. The genetic structure of this island is identical in L. monocytogenes and L. ivanovii, slightly different in L. seeligeri, and absent from all remaining species (Gouin et al., 1994). LIPI-1 is stably inserted at the same chromosomal position in all three species and there are no obvious traces of mobility genes, insertion sequences, or direct repeats which are also associated with horizontally transferred elements. In addition, its GC content does not differ significantly from L. monocytogenes chromosomes. This makes it difficult to confirm that this island was horizontally transferred but nonetheless it is categorized as a such due to the fact that is it inserted between prs and idh encoding  43  two housekeeping genes which are highly conserved in all Listeria spp. A number of studies have shown that prfA encoded on LIPI-1 co-regulates genes from the σB regulon demonstrating that this island has roles in both L. monocytogenes virulence and overall stress response. However, as it is found in all L. monocytogenes strains it is not a useful genetic determinant of stress tolerance. Similarly, LIPI-2 is a 22 kbp chromosomal locus that is uniquely specific to L. ivanovii and thus is also not a useful phenotype determinant for L. monocytogenes.   The third pathogenicity island, LIPI-3, was first described by Cotter et al. (2008) and is approximately 9 kbp in size and consists of 14 genes. The primary concern associated with this island is the presence of an eight-gene cluster encoding listeriolysin S which belongs to the extended family of streptolysin S-like peptides. These peptides are predominantly found within group A streptococci, and they significantly enhance the pathogenicity of carrier strains by contributing to cytotoxicity, inflammatory activation, and polymorphonuclear neutrophil (PMN) resistance, thereby playing a role in necrosis and systemic spread (Nizet et al., 2000). Cotter et al. (2008) demonstrated that listeriolysin S plays a role in the survival of L. monocytogenes in PMNs and contributes to its virulence in a mouse model. In the same study LIPI-3 was identified among 52% of LI strains (n=44) but was absent from all lineage II and LIII strains as well as all other Listeria spp. tested (L. innocua, L. welshimeri, L. seeligeri, L. ivanovii, and L. grayi). Furthermore, among the LIPI-3+ strains were six strains responsible for epidemic outbreaks. It is therefore suspected that this island may be the genetic basis for the enhanced virulence exhibited by a proportion of LI L. monocytogenes strains.  LGI1 is a 50 kbp genomic island that was first identified in Canadian CC8 isolates associated with a large 2008 listeriosis outbreak involving contaminated deli meats and resulting in 22 fatalities (Gilmour et al., 2010). Since then, LGI1 has been identified in other CC8 L.  44  monocytogenes isolates from Canada (Kovacevic et al., 2013) but not from other countries (Althaus et al., 2014). LGI1 is inserted between genes with 98% identity to lmo1702 and lmo1703 and shows evidence of horizontal transfer in that it contains serine recombinases and 16 bp imperfect inverted repeats at either end (Gilmour et al., 2010). The conservation of LGI1 among Canadian isolates and its association with a fatal outbreak has led to heightened interest in the putative functions of the genes located on this island which encode putative type II and type IV secretion systems, pilus-like surface structures, a multidrug efflux pump homologue (EmrE), and an alternative sigma factor (Gilmour et al., 2010). Kovacevic et al. (2015) reported that deletion of LGI1 genes with putative efflux (emrE), regulatory (lmo1852), and adhesion (sel1) functions, had no impact on the tolerance of L. monocytogenes to acid, cold, or salt, but that deletion of the effux pump emrE, increased susceptibility to quaternary ammonium-based sanitizers.  Recently a second Listeria genomic island (LGI2) has been discovered in L. monocytogenes and it spans ~35 kbp in some serotype 4 strains (4a, 4b, 4e) and is integrated into genes orthologous to lmo2224 (Briers et al., 2011; Kuenne et al., 2013; Lee et al., 2013). Encoded on this mobile element are 36 genes with roles in arsenic and cadmium resistance, and DNA integration, conjugation, and pathogenicity (internalin gene). LGI2 is putatively thought to have been inserted by means of a bacteriophage integrase (Briers et al., 2011; Kuenne et al., 2013). At this point the island is yet to be associated with enhanced virulence or stress tolerance but the encoded genes suggest that it is possible.     45  1.3.3.6.3 Bacteriophages Another common feature of L. monocytogenes genomes is the presence of prophages which are bacteriophage genomes that have integrated into a bacterial chromosome or plasmid. A comparative genomic study of four L. monocytogenes strains from three different serotypes observed the presence of one to four prophages per strain (Hain et al., 2012). Global transcriptomics of these strains during intracellular infection of host cells revealed that regardless of the strain lineage or prophage location, several prophage genes were upregulated in all strains (Hain et al., 2012). Furthermore, in the same study deletion of one of two prophage loci in L. monocytogenes EGD-e resulted in severe attenuation of virulence in a murine infection model demonstrating that prophages contribute to the pathogenicity of L. monocytogenes. Other studies have also shown that prophages can provide increased growth under nutrient limiting conditions (Edlin et al., 1977), can contribute to increased biofilm formation (Fortier and Sekulovic, 2013; Verghese et al., 2011; Wang et al., 2010) and can be beneficial for withstanding osmotic, oxidative, and acid stress (Wang et al., 2010).  Schmitz-Esser et al. (2015) observed that prophages were highly conserved among nine L. monocytogenes ST121 (serotype 1/2a) strains which are commonly found in food production environments. Similar to the findings of Hain et al. (2012), ST121 genomes harboured between 1-4 prophages and consistent with traditional genomic islands, the prophages were predominantly inserted between tRNA genes. In three strains prophages were integrated into the comK gene. Prophages integrated into comK have been suggested to be important for the adaptation of L. monocytogenes to food environments (Verghese et al., 2011). Schmitz-Esser et al. (2015) predicted that the presence and high conservation of L. monocytogenes ST121 prophages might thus be advantageous for surviving the stress conditions in food production environments. Prophages  46  commonly contain viral genes encoding tail, base, and head proteins, cell lysis and lysogeny control proteins, DNA packaging proteins, and DNA replication and recombination proteins (Verghese et al., 2011). However, they also contain a number of hypothetical proteins of unknown function. While the exact mechanisms relating prophage harbourage to bacterial stress tolerance remain unknown, these findings highlight that prophages may serve as valuable biomarkers for predicting both the stress tolerance and virulence potential of L. monocytogenes isolates. 1.3.3.6.4 Transposons  Transposons are horizontally transferred elements that are characterized by their ability to jump from one location to another within a genome or between genomes and plasmids. Unlike genomic islands which frequently insert at tRNA genes, transposons integrate relatively randomly in hosts [reviewed in Salyers et al. (1995)] and frequently carry genes for antibiotic or heavy metal resistance. These genes are typically flanked by an insertion sequence on either end that encode proteins required for transposition (Salyers et al., 1995).  In a comparative genomic study by Kuenne et al. (2013), four transposons were identified among 11 out of 16 genomic sequences from 12 L. monocytogenes serotypes. These transposons included Tn916 (one serotype 7 strain), Tn554 (two LII strains), and IS3-like transposons (11 LI and LII strains). Tn916 contains cadmium resistance genes, while Tn554 contains an arsenate resistance operon, and the IS3-like transposons contain several genes encoding cell surface-associated proteins. In another study by Schmitz-Esser et al. (2015), identical copies of a transposon termed Tn6188 were found in all 10 ST121 L. monocytogenes strains analyzed. Prior to this study Müller et al. (2013;2014) found that this transposon was prevalent in 11% of 1/2a strains analyzed and that it confers resistance to quaternary ammonium compounds. L.  47  monocytogenes plasmids can also contain transposons with Tn5422 being among one of the more studied ones. Like Tn916, this transposon contains cadmium resistance genes and accordingly has been shown to confer increased resistance to this heavy metal (Lebrun et al., 1994). L. monocytogenes does not typically harbour transposons containing antibiotic resistance genes; however, studies have shown that such transposons can be transferred to L. monocytogenes from other bacteria such as Enterococcus (Doucet-Populair et al., 1991). Currently it remains unknown whether L. monocytogenes transposons contribute to the survival of L. monocytogenes in foods but it is clear that some have roles in surviving in the natural environment as well as sanitary control measures. 1.3.3.7 Single nucleotide polymorphisms (SNPs) In addition to the presence or absence of genes, single nucleotide polymorphisms (SNPs) can also impact strain phenotypes. While each SNP represents a difference in a single nucleotide in a sequence, they are responsible for the majority of genetic variation among organisms. SNPs can result from spontaneous mutations that occur during DNA replication, or as a result of a selective pressure such as the presence of antibiotics (Martinez and Baquero, 2000). Generally speaking, bacterial chromosomes have a mutation rate of 1/300 per genome per replication (Drake et al., 1998). This mutation rate comprises all kinds of mutations including base pair substitutions, or base insertions and deletions which can cause frameshifts in coding regions (Drake et al., 1998).  Depending on where a SNP occurs, it can have different consequences at the phenotypic level. SNPs in coding regions can be synonymous resulting in the same amino acid being encoded, or non-synonymous which can result in a new amino acid being encoded or in the case of a nonsense mutation, a premature stop codon (Syvänen, 2001). SNPs in regulatory regions may  48  affect the attachment of RNA polymerases and transcription factors leading to increased or decreased operon expression. Overall, the rate of disadvantageous mutations per genome per replication is known to be much higher (~2–8 × 10−4 in E. coli) (Boe et al., 2000; Kibota and Lynch, 1996), than that of beneficial mutations (~2 × 10−9 in E. coli) (Imhof and Schlotterer, 2001) highlighting that while mutations associated with enhanced stress tolerance do occur, they are more difficult to discover.  Nelson et al. (2004) compared the genomes of three L. monocytogenes strains associated with food-borne illness in the US and found that strain H7858 (4b) and strain F6854 (1/2a) contained 8603 and 105,050 high quality SNPs respectively, compared to F2365 (4b) which was selected as the reference. Of these SNPs, 16-23% resulted in non-synonymous changes and were commonly found in genes with roles in energy metabolism and transport. In a similar study by Hain et al. (2012), surface proteins showed the highest number of SNPs between two closely related 4b genomes and between a 4a genome and EGD-e (1/2a). Specifically, genes encoding LPXTG-motif proteins appeared to be the most affected. As discussed earlier, many of the phenotypic differences between L. monocytogenes serotypes and strains are believed to be a result of diversity among cell surface proteins.  While we now have the tools to detect SNPs easily amongst bacterial species, the sheer number of SNPs between strains makes it difficult to link a single mutation to a particular phenotype. To date, studies that have been successful at making these associations have studied spontaneous or artificial mutations stemming from a single strain. For example, in the study by Hoffmann et al. (2013), an artificially produced thymine deletion in the σA-like promoter region of betL, encoding an osmolyte (betaine) transporter, dramatically increased betL transcription, and resulted in enhanced osmo- and chill-tolerance in L. monocytogenes. However, this mutation is yet  49  to be observed among any naturally occurring strains. In contrast, Karatzas et al. (2003) reported that a spontaneous high hydrostatic pressure tolerant L. monocytogenes mutant of Scott A, contained a single codon deletion in ctsR, a negative regulator of several heat shock and general stress proteins, that also conferred increased thermo-tolerance and resistance to H2O2. Additionally, Metselaar et al. (2015) observed a number of L. monocytogenes spontaneous acid tolerant mutants contained SNPs in the ribosomal protein gene rpsU. Ribosomal proteins are known to assist bacterial adaptation to certain stresses including cold, osmotic, heat, and oxidative stress (Durack et al., 2013; Gardan et al., 2003; Hecker and Völker, 1998; Völker et al., 1994) Perhaps one of the most well-known set of SNPs in L. monocytogenes are those responsible for the various PMSCs observed in inlA. To date, 19 different inlA PMSCs have been identified (Hu et al., 2007a; Jonquieres et al., 1998; Nightingale et al., 2008, 2005a, Olier et al., 2003, 2002; Rousseaux et al., 2004; Van Stelten et al., 2010; Van Stelten and Nightingale, 2008; Wu et al., 2016). As discussed earlier, these result in truncated versions of the protein that are secreted from the cell as opposed to being attached, and are associated with attenuated virulence. Kovacevic et al. (2013) recently discovered that full-length variants of inlA were more prevalent among fast cold-adapting L. monocytogenes strains compared to intermediate and slow cold-adapting strains, suggesting that inlA profiling may also be suitable for predicting the cold tolerance of strains. Other PMSCs have also been reported in L. monocytogenes stress response genes including in prfA and sigL (Handa-Miya et al., 2007; Nightingale et al., 2007). However, aside from reduced virulence these truncations have not been associated with any stress-tolerance phenotypes.    50  1.3.3.8 Regulatory RNAs Small non-coding regulatory RNAs (ncRNA) represent a newly discovered layer of gene expression regulation in prokaryotes. The development of more advanced molecular techniques, has allowed scientists to unveil the abundance of these molecules within genomes as well as their regulatory roles. ncRNAs regulate transcription by pairing with other RNAs, forming parts of RNA-protein complexes, or adopting regulatory secondary structures such as riboswitches (Storz et al., 2004). To date, over 200 ncRNAs have been identified in L. monocytogenes and been implicated in responses to iron limitation, oxidative stress, and intracellular growth (Christiansen et al., 2006; Mraheil et al., 2011; Mandin et al., 2007; Nielsen et al., 2009; Ponting et al., 2009; Toledo-Arana et al., 2009; Wurtzel et al., 2012). Kuenne et al. (2013) screened 11 L. monocytogenes strain genomes (representing six serotypes) for 210 ncRNAs and found that 43 fell into the category of accessory ncRNAs, defined as being absent from at least one compared strain. This included 20 ncRNAs that were found to only be present in a subset of lineage II strains, and have been previously suggested to be involved in virulence or pathogenicity. One of these ncRNAs was found to be inserted as part of prophage A118 in a strain which may in part explain why prophages are induced in L. monocytogenes under certain conditions (Hain et al., 2012). While there is still much to learn regarding the roles of ncRNAs, these preliminary findings suggest that the presence-absence of certain ncRNAs within genomes likely impact strain phenotypes.  1.3.4 L. monocytogenes stress response mechanisms Bacterial stress from a food microbiology perspective is considered to be a physical, chemical, or nutritional condition insufficiently severe to kill, resulting in sublethally injured microbes (Wesche et al., 2009). Stress induced challenges include membrane transitional states,  51  starvation, protein denaturation, and structural changes to nucleic acid structures as reviewed in Wesche et al. (2009). The following sections will discuss some of the response mechanisms specifically employed by L. monocytogenes to overcome these challenges.  1.3.4.1 Membrane lipid compositional changes L. monocytogenes and bacteria at large, are known to alter their membrane lipid composition in response to almost all forms of physical stress (Russell et al., 1995). Such changes are needed in order to maintain membrane integrity and transport functions which are required for cell survival. However, since each bacterium possesses slightly different manufacturing machinery, stress induced membrane lipid alterations differ between bacterial strains, species, genera, and phyla (Mazzotta and Montville, 1997; Russell et al., 1995).  The L. monocytogenes lipid membrane consists predominantly of branched-chain fatty acids (BCFAs, ~80-90%). The four most abundant BCFAs, listed in decreasing order, are anteiso-C15:0 (a-C15:0), a-C17:0, iso-C15:0 (i-C15:0) and i-C17:0 (Annous et al., 1997; Juneja et al., 1998; Verheul et al., 1997; Zhu et al., 2005). The remaining fatty acid proportion consists of saturated (SFAs) and unsaturated fatty acids (UFAs) ranging from 12 to 20 carbons in length depending on the growth conditions.  When subjected to temperature stress, L. monocytogenes modifies its membrane to either increase or decrease the phase transition temperature such that it is able to maintain a proper level of fluidity for membrane integrity and transport functions (McElhaney, 1982). If a membrane is too fluid-like it will become leaky, permitting important molecules to exit the cell and likewise unfavorable molecules to enter. Similarly, if a membrane is not fluid enough, it will enter a liquid-crystalline state which frequently disrupts membrane transport (McElhaney, 1982). While some  52  minor changes have been noted in the membrane composition of L. innocua following 1 h of exposure to 45°C compared to growth at 22°C (Moorman et al., 2008), most Listeria membrane-composition studies have focused on the changes that occur during growth at lower temperatures.  When L. monocytogenes experiences a decrease in temperature, it increases its relative membrane proportion of a-C15:0 by ~20% in cells downshifted from 20-30°C to 5-10°C, and correspondingly, decreases its levels of a-C17:0 by the same amount (Kaneda 1991; Zhu et al., 2005). This is favourable because a-15:0 has a melting point of 24.0°C compared to 36.8°C for a-C17:0, thus increasing membrane fluidity (Kaneda, 1991). Similarly, Listeria also incorporates FAs with shorter chain lengths as well as those with a higher degree of unsaturation into its membrane to further decrease the phase-transition temperature when grown at low temperatures (Moorman et al., 2008; Zhu et al., 2005).  In response to acid stress (pH 5.5, 1 h) L. monocytogenes has been shown to increase its levels of the SFAs C14:0 and C16:0, and decrease levels of C18:0 (van Schaik et al., 1999). In a similar study by Giotis et al. (2007), L. monocytogenes grown at pH 5.5 or 6.0 was found to have higher proportions of SFAs, while at pH 8.0 and 8.5 it contained higher proportions of BCFAs, specifically those with an anteiso formation. Juneja et al. (1998) examined the ratio of C15:0 to C17:0 in L. monocytogenes grown at pH 5.4 vs. pH 7.0 and found it was only dependent on growth temperature and not the type of acidulant or the pH. In all three studies the changes that occurred in response to pH stress were very minor (2-5%) compared to those which have been reported to occur in response to low temperature stress (20-25%). This is likely due to the fact that unlike temperature stress, pH stress affects the overall charge on membranes and such conditions are instead usually associated with changes in the head-groups of lipids. For example, in response to salt stress, the largest membrane alteration observed in bacteria is an increase in the proportion of  53  anionic phospholipids and/or glycolipids (Romantsov et al., 2009; Russell, 1995). When L. monocytogenes is grown in the presence of 2% NaCl it increases its ratio of diphosphatidylglycerol/phosphatidylglycerol (Russell et al., 1995). Specifically, cardiolipin (a diphosphatidylglycerol) is known to increase in bacteria in response to salt stress and is thought to play a role in the regulation of osmolyte transporters (Romantsov et al., 2009).  1.3.4.2 Regulatory elements 1.3.4.2.1 RNA polymerase sigma factors Sigma factors are bacterial transcription initiation factors that enable specific binding of RNA polymerase to gene promoters. L. monocytogenes genomes are known to encode four alternative sigma factors (σB, σC, σH, σL) and one principal RNA polymerase sigma factor (RpoD) which regulates housekeeping genes associated with ribosome structure, protein synthesis, and rRNA and tRNA synthesis and processing (Metzger et al., 1994). Of the four alternative sigma factors, σB has been the most extensively studied and is known to positively regulate over 150 genes when the organism enters stationary phase or is subjected to environmental stresses including low pH, high salt, or carbon starvation (Becker et al., 1998; Ferreira et al., 2001; Fraser et al., 2003; Kazmierczak et al., 2003; Wiedmann et al., 1998). The alternative sigma factor σL is also known to play a large role in the L. monocytogenes stress response and has been shown to contribute to the pathogens’ ability to tolerate cold, osmotic, and acid stress (Okada et al., 2006; Raimann et al., 2009). This sigma factor positively regulates >400 genes, including those involved in cell envelope synthesis, motility, and phosphotransferase system (PTS) sugar uptake and catabolism (Arous et al., 2004; Mattila et al., 2012). Lastly, the remaining alternative sigma factors σH and σC, have been less studied but are known to contribute to the survival of L. monocytogenes  54  under alkaline and heat stress, respectively (Phan‐Thanh and Mahouin, 1999; Rea et al., 2004; Zhang et al., 2005). 1.3.4.2.2 Global regulatory proteins In addition to sigma factors, L. monocytogenes genomes also encode a number of both positive (PrfA) and negative (HrcA, CtsR, CodY, MogR, CtsR, HrcA) global regulatory proteins. CodY is a pleiotropic transcriptional regulator that actively represses the transcription of genes involved in amino acid metabolism, nitrogen assimilation, mobility and chemotaxis, and sugar uptake, among others (Bennett et al., 2007). DegU is another well-known response regulator that regulates the expression of motility-, virulence-, and biofilm-related genes in L. monocytogenes (Gueriri et al., 2008; Knudsen et al., 2004; Williams et al., 2005). The transcriptional repressor MogR also controls motility by repressing flagella motility genes at high temperature (37°C) (Gründling et al., 2004). HrcA and CtsR are two additional negative regulators that predominantly repress genes involved in the heat-shock response and DNA and protein quality control (e.g. dnaK, grpE, groES, groEL, clpB, clpC, clpE, clpP) (Hanawa et al., 2000; Karatzas et al., 2003; Nair et al., 2000). Lastly, PrfA positively controls the virulence regulon in L. monocytogenes; however, many of these genes also have roles in stress tolerance and accordingly, are co-regulated by σB (Chakraborty et al., 1992).  1.3.4.2.3 Two-component systems Two-component signaling systems (TCSs) are also important regulators of bacterial stress responses and can sense changes in the environment including low pH, and osmotic, oxidative, and ethanol stress (Bourret, 2010; Kallipolitis and Ingmer, 2001). TCSs consist of a membrane-associated histidine kinase (HK), and a response regulator (RR) which enables the cell to respond  55  by altering gene expression (Cotter et al., 1999). The sequenced genome of L. monocytogenes EGD-e contains 16 known TCSs (Glaser et al., 2001). Two of the more well-characterized TCSs are LisRK and KdpE. LisRK consists of lisR and lisK which encode the RR and HK, respectively. This system has been shown to regulate virulence and responses to low pH (Cotter et al., 1999). KdpE on the other hand is an uptake system involved in the inward transport of potassium (K+) in response to osmotic stress (Kallipolitis and Ingmer, 2001). Two additional two-component regulatory systems in L. monocytogenes; CheAY and YycGF, are involved in regulating chemotaxis (Dons et al., 2004) and cell wall modifications (Dubrac and Msadek, 2008), respectively.  1.3.4.2.4 Non-coding RNA As previously discussed, L. monocytogenes genomes have been recently shown to encode a large pool of ncRNAs with some being strain or serotype specific while others are conserved among all L. monocytogenes strains. Compared to regulatory proteins, ncRNA transcripts are able to provide a more rapid means of communication between quorum sensing and destabilization of target mRNAs, thereby allowing bacteria to more tightly control protein expressin under environmental stress conditions (Izar et al., 2011; Thomason and Storz, 2010). The majority of ncRNAs have been shown to regulate the synthesis of transcription regulators, in particular alternative sigma factors associated with stress responses [reviewed in Repoila and Darfeuille (2009)]. For example, Nielsen et al. (2008) discovered a novel small ncRNA in L. monocytogenes that is produced in a σB dependent manner.  To date, the expression profiles of ncRNAs in L. monocytogenes have predominantly been evaluated using murine models or host intracellular conditions (Behren et al., 2014; Toledo-Arana  56  et al., 2009; Wehner et al., 2014; Wurtzel et al., 2012). However, in other bacteria ncRNAs have been shown to become activated in response to temperature, nutrient, oxidative, pH, anaerobic, and iron deficiency stresses [reviewed in Hoe et al. (2013)] making it reasonable to expect that they contribute to similar stress responses in L. monocytogenes as well.  1.3.4.3 Stress induced proteins 1.3.4.3.1 Osmolyte and oligopeptide transporters When L. monocytogenes is subjected to low temperature, high osmolarity, acid, or heat stress, it is known to activate genes responsible for the intracellular accumulation of solutes and short peptides (Sleator et al., 2001; van der Veen et al., 2007; Wemekamp-Kamphuis et al., 2004a). These solutes and peptides function as osmoprotectants that have been shown to facilitate cell growth under stressful conditions by maintaining or enhancing protein stability (Ali et al., 2017; Wesche et al., 2009). In L. monocytogenes, carnitine and glycine betaine are the predominant solutes that accumulate, and proline uptake is used to compensate when the other two osmolytes are not available (Sleator et al., 2003). These uptake systems are encoded by the opuCABCD, gbuABC, and proBA operons, respectively. Short peptides on the other hand, are taken up by L. monocytogenes via the membrane permease OppA (oppABCDF) which can transport peptides containing up to eight residues (Borezee et al., 2000). Specifically, di- and tripeptides containing glycine and proline are responsible for stimulating the growth of L. monocytogenes under osmotic and cold stress conditions (Mario-Rosario et al., 1995).    57  1.3.4.3.2 Cold shock proteins Cold stress or cold shock proteins (CSPs) are a conserved family of small (~70 aa) proteins containing a nucleic acid-binding domain. Found in many prokaryotic and eukaryotic organisms, CSPs can act as transcriptional activators, antiterminators, or as RNA chaperones which enhance translation at low temperatures by blocking the development of secondary mRNA structures (Hunger et al., 2006; Jiang et al., 1997; Phadtare, 2004). Three CSPs have been identified in L. monocytogenes and are listed here in the order of functional importance: CspL>CspD>CspB (Schmid et al., 2009). These proteins have been shown to only be induced in the early stages following cold stress (Chan et al., 2007b; Durack et al., 2013; Schmid et al., 2009). Despite the demonstrated roles of CSPs in the L. monocytogenes cold-stress response, other studies have shown that they are also upregulated in response to osmotic stress (Schmid et al., 2009) and high hydrostatic pressure (Wemekamp-Kamphuis et al., 2002).  1.3.4.3.3 Heat shock proteins Similar to CSPs, L. monocytogenes also induces the transcription of several heat shock proteins (HSPs) in response to more than just heat stress. These proteins belong to one of two classes: molecular chaperone proteins or adenosine triphosphate (ATP)- dependent proteases (ATPases) which aid in refolding and breaking down damaged proteins, respectively (Hu et al., 2007a). HSPs identified in L. monocytogenes include GroES, GroEL, GpmA, DnaK, DnaJ, HtrA, McsAB, and several Clp proteins (Stack et al., 2008). Many of these genes are positively regulated by σB and negatively regulated by the transcription repressors HrcA and CstR (Stack et al., 2008). In addition to heat stress, increased expression of many of these genes have been reported in  58  response to salt, cold, intracellular survival, and ethanol stress (Kilstrup et al., 1997; Liu et al., 2002; Salotra et al., 1995).  1.3.4.3.4 RNA and DNA repair proteins Nucleic acids are easily damaged and RNA and DNA breaks accumulate under oxidative stress conditions which frequently occur alongside other stresses (Mittler, 2002). Furthermore, at low temperatures, both DNA replication and transcription are hindered by cold-induced changes in nucleic acid structures. In response to these challenges, L. monocytogenes is known to upregulate genes encoding topoisomerases and DNA gyrases, which help maintain the super helical tension of DNA; RNA helicases, which unwind secondary RNA structures; and exo- and endonucleases, which function in nucleic acid repair (Chan et al., 2007b; Durack et al., 2013; Markkula et al., 2012).   59  Chapter 2: Genotypes associated with Listeria monocytogenes isolates displaying impaired or enhanced tolerances to cold, salt, acid, or desiccation stress  2.1  Introduction Listeria monocytogenes is a ubiquitous bacterial foodborne pathogen that is most recognized for its ability to grow at temperatures as low as -0.4°C (Walker et al., 1990) and cause listeriosis, a serious disease with an average mortality rate of 30% among at-risk people (Yildiz et al., 2007). In addition to possessing cold tolerance, L. monocytogenes is also capable of surviving many other food-related stresses including high osmolarity (Shabala et al., 2008) and low pH (Sorrells et al., 1989), further adding to its hardiness. Additionally, cross contamination of foods is facilitated by biofilm formation (Chavant et al., 2002; Di Bonaventura et al., 2008; Hingston et al., 2013; Moltz, 2005; Møretrø and Langsrud, 2004), and the ability of the organism to survive desiccation for extended periods of time on food contact surfaces (Vogel et al., 2010). Post-processing levels of L. monocytogenes contamination in foods are usually low (Cabedo et al., 2008; Fenlon et al., 1996; Kozak et al., 1996) and unlikely to cause disease (Buchanan et al., 1997; Chen et al., 2003). It is therefore, refrigerated, ready-to-eat (RTE) foods with extended shelf lives and the potential for regrowth that present the largest risk to consumers. Both Canada and the EU have adopted regulations for the control of L. monocytogenes in RTE-foods, (Health Canada, 2016; Luber, 2011), allowing up to 100 CFU/g in foods that do not permit growth beyond this level within the shelf-life of the product, and a zero tolerance policy for foods identified as supporting growth. When validating growth inhibition of L. monocytogenes in  60  stabilized RTE-foods, it is important that the strains used represent the extremes of the behavior that has been observed among L. monocytogenes strains. In the US, the zero tolerance policy is applicable for all food products (US FDA, 2015). However, nationwide outbreaks continue to occur in the US. To date, there have been three multistate listeriosis outbreaks in 2016 that were associated with frozen vegetables, packaged salads, and raw milk and resulted in 20 illnesses and five deaths (US CDC, 2017). These numbers are yet to exceed that of 2015 where two multistate outbreaks involving soft cheese and ice cream resulted in 40 illnesses and six deaths (US CDC, 2017).  In 2013, the US established the Listeria Whole Genome Sequencing (WGS) Project to assist in detecting, investigating, and mitigating foodborne outbreaks (US CDC, 2016). Though valuable for tracing outbreaks, WGS is not routinely used to determine the stress tolerance of outbreak strains. However, WGS provides the information that could potentially lead to identification of molecular biomarkers related to the stress tolerance of L. monocytogenes isolates. Such biomarkers could greatly aid in monitoring the risks of L. monocytogenes contamination and regrowth in food products and processing environments (Jacquet et al., 2004).  Currently, one common molecular biomarker used for the virulence potential of L. monocytogenes is the internalin A encoding gene (inlA) which can contain one of several different premature stop codons (PMSCs) producing truncated and secreted proteins associated with attenuated virulence (Felicio et al., 2007; Handa-Miya et al., 2007; Jacquet et al., 2004; Jonquieres et al., 1998; Nightingale et al., 2005a; Roldgaard et al., 2009; Rousseaux et al., 2004; Van Stelten et al., 2011). Kovacevic et al. (2013) discovered that full-length variants of inlA were more prevalent among fast cold-adapting L. monocytogenes strains than intermediate and slow cold-adapting strains, suggesting that inlA profiling may also be suitable for predicting the cold  61  tolerance of strains. Another potential biomarker is the L. monocytogenes stress survival islet 1 (SSI-1). Included in this five gene cluster (lmo0444 – lmo0448) are two genes (gadT1 and gadD1) from the glutamate decarboxylase acid resistance system which has been shown to significantly improve the growth of L. monocytogenes in mildly acidic environments (Cotter et al., 2005). Additionally, an L. monocytogenes SSI-1 deletion mutant exhibited impaired growth at low pH (pH 4.8), high salt (7.5% NaCl), and on frankfurters stored at 4°C (Ryan et al., 2010). Further research with naturally occurring SSI-1 positive and negative strains is needed to determine if this island would be a suitable biomarker for predicting stress-tolerance phenotypes.  To date, studies that evaluated the stress tolerances of L. monocytogenes isolates have focused on associating phenotypes with genetic lineages (Bergholz et al., 2010), serotypes (Barbosa et al., 1994; Junttila et al., 1988; Ribeiro et al., 2014), and isolation sources (Begot et al., 1997; Durack et al., 2013). However, few significant differences between these groups have been observed, suggesting that the diversity among isolates within these means of classification is not definitive for predicting phenotypic behaviour. Instead, stronger phenotype associations might be observed among more closely related isolates, e.g., those sharing the same sequence type (ST) or clonal complex (CC). Additionally, the presence of specific genetic elements (e.g., inlA and SSI-1) may also influence the stress-tolerance phenotypes of isolates as well as more minor genetic differences such as single nucleotide polymorphisms (SNPs).   The objective of this study was to use a combination of phenotypic analyses and WGS to elucidate novel associations between L. monocytogenes genotypes and food-related stress-tolerance phenotypes with the goal of identifying biomarkers that can be used to predict the stress tolerances of food-chain isolates. To accomplish this, 166 L. monocytogenes isolates were evaluated on their ability to grow in cold (4°C), salt (6% NaCl), and acid (pH 5) stress conditions  62  as well as survive desiccation stress (33% RH). Factors investigated for potential associations with the observed phenotypes were: genetic lineage, serotype, CC, inlA profiles, and the presence of a plasmid, SSI-1, unique SNPs, and Listeria genomic island 1 (LGI1). 2.2 Materials and methods 2.2.1 Isolates and culture conditions A collection of 166 L. monocytogenes isolates from Canada and Switzerland were used in this collaborative study. This included: (i) 159 food and food-processing environment isolates from Canada (n=139) and Switzerland (n=20) (ii) six isolates from sporadic human listeriosis cases in Switzerland and (iii) one isolate from an asymptomatic human (Supplementary Table 2-1). All human isolates were anonymized and no ethical approval was required as per the institutional and national guidelines. Isolates were stored at -80°C in brain heart infusion broth (BHIB, Difco, Fisher Scientific, Canada) + 20% glycerol and routinely cultured at 30°C on BHI agar (Difco, Fisher Scientific) plates. 2.2.2 Whole genome sequencing Genomic DNA was isolated using the PureLink Mini Kit from Life Technologies, Canada. PicoGreen quantification was performed (Invitrogen, Canada) and DNA quality and quantity was determined using the NanoDrop 2000 (Fisher Scientific). Genomic DNA samples of sufficient quality and quantity were sequenced by Genome Quebec (Montréal, QC, Canada) using TruSeq automated library preparation (Illumina) and paired-end, 100 bp sequencing on the Illumina Hi-Seq. Between 4.9-16.5 million high quality reads remained after quality control for each genomic library. Raw FASTQ files were trimmed using Cutadapt in Trim Galore! version 0.4.1 and de novo genome assembly was performed using SPAdes version 3.1.0 (careful option used) (Bankevich et  63  al., 2012). Low coverage (<10) and small contigs (< 200 bp) were removed from assemblies using a custom perl script. Assemblies were subsequently annotated using Prokka version 1.5.2 (genus Listeria, species monocytogenes) (Seemann, 2014). Assembled sequences were deposited into the NCBI Whole Genome Shotgun (WGS) database under Bioproject PRJNA329415. 2.2.3 Lineage determination To classify isolates into genetic lineages, a reference free, k-mer based SNP phylogeny was generated using the kSNP 3.0 program (Gardner et al., 2015) and reference isolates for the major lineages of L. monocytogenes (LI - F2365; LII – EGD-e; LIII – HCC23; LIV - J1-208). The resulting maximum parsimony tree (based on the consensus of 100 trees) clearly segregated the four lineages. 2.2.4 Multi locus sequence typing To group isolates based on their epidemiological context, in silico MLST was performed using the Center for Genomic Epidemiology’s MLST typing tool (https://cge.cbs.dtu.dk/services/MLST/). Clonal Complexes (CCs) were assigned based on the Pasteur Institute schema (http://www.pasteur.fr/recherche/genopole/PF8/mlst/Lmono.html). Novel sequence types (STs) were confirmed using Sanger sequencing and submitted to the Pasteur Institute Database for new assignments (http://bigsdb.web.pasteur.fr/Listeria/Listeria.html).  2.2.5 In silico serogroup/serotype assignment Antibody-based serotyping was conducted on a subset of isolates (n=91) within both the current study and previous studies (Arguedas-Villa et al., 2010; Kovacevic et al., 2013). Remaining isolates were assigned one or more possible serotypes by performing a MegaBLAST  64  search (>95% nt identity) ncbi-blast+ v. 2.3.0 available at: ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST/) for four genes used in a multiplex PCR developed by Doumith et al. (2004). Additionally, predictive serotypes were assigned to isolates with STs that are known to be associated with a specific serotype. 2.2.6 Targeted genomic element screenings The genes and genomic regions evaluated in this study were 1) the plasmid replicon gene repA, used to indicate the presence of a plasmid (Kuenne et al., 2010), 2) emrE, representing Listeria genomic island 1 (LGI1), a 50 kbp island with putative roles in stress tolerance and persistence (Gilmour et al., 2010), and 3) SSI-1, a five gene cluster previously identified as having a role in the response of L. monocytogenes to cold, osmotic, and acid stress conditions (Ryan et al., 2010). Additionally, the coding sequence of inlA was investigated to determine if isolates possessed a full-length sequence or PMSC mutation. emrE, SSI-1, and inlA were screened for among isolate sequence assemblies using MegaBLAST (>95% nt identity) and repA was screened for using BLASTP (>30% aa identity over >80% coverage). inlA and repA sequences were then extracted from the isolate assemblies for further analysis.   2.2.7 Identification of putative plasmid contigs  Detection of repA sequences meant that at least one contig belonged to a putative plasmid. To identify additional plasmid associated contigs, isolate assemblies were aligned to the closed genome of L. monocytogenes EDG-e (Accession: NC_003210.1) using Contig Mover in Mauve version 2.3.1 (Rissman et al., 2009). Contigs not aligning to the EDG-e chromosome were compared to published L. monocytogenes plasmids (Kuenne et al., 2010) by BLAST. Contigs were excluded if they displayed open readings frames associated with chromosomal DNA (e.g. rRNA,  65  tRNA) or did not align to any of the Listeria associated plasmids annotated by Kuenne et al. (2010): pLM33, pLM1-2bUG1, pLM5578, pLM80, and pLI100. A summary of the putative plasmid contigs found within each isolate can be found in Supplementary Table 2-1. 2.2.8 Cold tolerance assay Overnight cultures grown in BHIB at 37°C were standardized to 109 CFU/ml using spectrophotometric methods, and diluted in pre-chilled BHIB to yield a final density of 103 CFU/ml and stored at 4°C [previously described in Arguedas-Villa et al., (2010)]. The bacterial density was enumerated daily for the first four days and then bi-weekly for up to five weeks by plating on tryptic soy agar (TSA, BD, Fisher Scientific) + 6% yeast extract (BD, Fisher Scientific). The resulting growth curves were fitted using a four parameter logistic model described by Dalgaard and Koutsoumanis (2001).  2.2.9 Salt and acid tolerance assay Isolates were assessed for salt and acid tolerance using modified versions of published protocols (Bergholz et al., 2010; Cotter et al., 2005; van der Veen et al., 2008). In short, overnight cultures grown in BHIB at 30°C were diluted in either BHIB+6% (w/w) NaCl or BHIB adjusted to pH 5 (with 1 M HCl) to achieve a final concentration of 107 CFU/ml. From these cultures, 200 µl was added in duplicate (technical replicates) to 96-well plates (Costar™ clear polystyrene, Fisher Scientific) that were incubated at 25°C in a microplate reader (Spectramax, V6.3; Molecular Devices, Sunnyvale, CA). A temperature of 25°C was used to assess isolate salt and acid tolerance under non-intracellular or cold stress conditions. The absorbance (A600nm) of each well was recorded every 30 min until all isolates reached stationary phase (~26 h) and the resulting growth  66  curves were fitted to the Baranyi and Roberts model (Baranyi and Roberts, 1994) using DMfit (v3.5) available on the ComBase browser (http://browser.combase.cc/DMFit.aspx).  2.2.10 Desiccation tolerance assay Cultures grown for 24 h in BHIB at 20°C were diluted to 107 CFU/ml in buffered peptone water (BD, Fisher Scientific) and 10 µl (105 CFU) was spotted in duplicate (technical replicates) on the bottom of wells in lid-less 96-well plates. The plates were then stored for three days at 20°C in desiccators (SICCO, Bohlender, Germany) pre-conditioned to 33% relative humidity (RH) using a saturated solution of MgCl2 [protocol adapted from Hingston et al. (2015)]. The RH of the chambers was monitored throughout the desiccation periods using data loggers (included with desiccators). A temperature of 20°C was used to simulate desiccation conditions that might occur in a food plant. Following desiccation, the plates were rehydrated with 200 µl of BHIB, and incubated at 25°C in a plate reader where the A600nm of each well was recorded every 30 min until all isolates reached stationary phase (~24 h). The resulting growth curves were then fitted to the Baranyi and Roberts model and the model parameters recorded.  2.2.11 Phenotype designations and statistical analyses For all four stress exposure experiments, a minimum of two biological replicates with two technical replicates each, were conducted for all isolates. Based on the findings of Aryani et al. (2015), the data was standardized for biological variability between replicates by dividing isolate growth parameters by the median value for each experimental run, thereby making the median equal to 1. The median was selected for standardization rather than the mean to avoid the influence of very stress sensitive isolates. Model parameters (LPD – lag phase duration; µmax – maximum growth rate; Nmax – maximum cell density) were averaged across biological replicates and  67  presented as standardized (std) values. For isolates where the average std values had a standard deviation (SD) >0.05, additional replicates were completed to obtain more representative means. Isolates were considered tolerant or sensitive to cold, salt, or acid stress if they had an average std-µmax > or < than 1SD from the median, respectively. All remaining isolates were considered to have intermediate stress tolerance. For desiccation stress survival the model parameter of most interest was the LPD, indicating the time to (detectable) regrowth (TRG) which is negatively correlated with the number of cells that survived the desiccation treatment. Isolates were classified as desiccation tolerant or sensitive if they had an average std-TRG < or > than 1SD from the median, respectively. A standard curve generated using five cell levels (101-105 CFU) produced a correlation of y=-0.25x+2.07 (R2=0.97) where y is the TDR and x is the log10 number of viable cells in each well following desiccation.  To elucidate potential associations between the factors we investigated, statistical tests were performed using IBM SPSS Statistics version 23. Specifically, individual two-tailed T-tests and one-way ANOVAs with Tukey post hoc tests were used to compare the average standardized stress tolerance model parameters of two (±plasmid, ±SSI-1, ±LGI1, lineage I and II, repA group 1 vs. group 2 isolates) or more groups (serotypes, CCs, inlA profiles, and sensitive, intermediate and stress tolerant groups), respectively. G*Power 3.1.9.2 (Faul et al., 2009, 2007) was used to determine the minimum sample sizes required to ensure a power of 0.80 for all statistical tests. All data sets were accessed for outliers, homogeneity of variances (Levene’s test), and normality (Shapiro-Wilk’s test). Where homogeneity of variances could not be achieved, Welch T-tests and Welch ANOVAs in combination with Games-Howell post hoc tests were used. P-values below 0.05 were considered significant for all comparisons.  68  2.2.12 Phylogenetic reconstruction based on core genome single nucleotide variants Parsnp, a tool within Harvest suite of tools (Treangen et al., 2014), was used to perform core genome alignment of all 166 de novo assembled genomes and the reference L. monocytogenes EGD-e strain in order to identify SNPs within the core genome. SNPs clustered within 20 base pairs were removed as these may indicate repetitive regions containing more erroneous SNP calls. The remaining high quality SNPs were used to generate maximum likelihood trees using the RaxML version 8 (Stamatakis, 2014) on the CIPRES science gateway (Miller et al., 2010) using default parameters (including the GTRCAT nucleotide model and 100 bootstrap replicates). Corresponding heatmaps containing additional genotype and phenotype information were generated in R version 2.15.1 (R Core Team, 2016) using the heatmap.2 function from the gplots library. 2.2.13 SNP detection SNPs were also detected against the L. monocytogenes EGD-e (NC_003210.1) reference genome. SMALT version 0.7.6 (http://www.sanger.ac.uk/science/tools/smalt-0) with default parameters except “-i 330” was used to first align raw reads against the reference. Samtools version 1.2 (Li, 2011) was used on these assemblies to sort the aligned reads (“samtools sort”), remove potential PCR duplicates (“samtools rmdup”) and call the SNPs (“samtools mpileup”). Additional filtering of SNP calls included removing those with a read depth less than 50 and heterozygous genotypes (since our genomes are haploid) using the “bcftools filter” command. SNPs found in repetitive regions of the genome as assessed by the index of repetitiveness (Schwender et al., 2004) were also removed manually. The remaining high confidence SNPs were annotated using SNPEff version 4.1 (Cingolani et al., 2012) with the Listeria_monocytogenes_EGD_e_uid61583  69  annotation. Synonymous SNPs were also removed in the end for identification of non-synonymous or potential regulatory SNPs that may be contributing to phenotypic differences in cold growth. 2.2.14 Statistical methods for elucidating SNPs associated with stress-tolerance phenotypes SNPSift version 4.1 “CaseControl” (Cingolani et al., 2012) was used to run a Fisher Exact test to identify SNPs that were significantly associated with case versus control groups. To identify SNPs only found in tolerant isolates, these were used as the case group, while all others were used as the control group. Since this did not yield any results, subsequently, the sensitive isolates alone were used as the control group to allow SNPs to be seen in intermediate growers. This method has certain limitations in that certain associations may require very large sample sizes to become statistically significant, especially considering genetic heterogeneity leads to the same phenotype or the potential for multiple SNPs to interact. An alternative approach, Random Forests™ (Breiman, 2001), was also used to discover important SNPs in distinguishing stress tolerant and sensitive groups since previous genome wide association studies have shown random forests outperform the Fisher Exact test in these special cases (Bulinski et al., 2011; Bureau et al., 2005; De Lobel et al., 2010; Lunetta et al., 2004; Schwender et al., 2004). The RandomForest™ version 4.6-10 library was used in R with default parameters except “importance=TRUE, proximity=TRUE, ntree=5000”. This allows the method to run as a classifier that then ranks SNPs on their ability to classify isolates based on their phenotypic designation. 2.2.15 Genomic islands analysis Annotated draft genomes were submitted to IslandViewer 3 (Dhillon et al., 2015) using L. monocytogenes EGD-e (NC_003210.1) as a reference for contig reordering. Genomic islands were predicted using IslandPath-DIMOB (Hsiao et al., 2005) and SIGI-HMM (Waack et al., 2006).  70  Predicted genomic islands positioned on the genome within less than 10 kbp of each other were merged into one single region. To form groups of similar genomic islands, the genetic distance between genomic island sequences was computing using Mash (parameter -s 2000) (Ondov et al., 2016) and groups of similar sequences were identified using hclust and cutree in R.  2.3 Results  2.3.1 Genetic characteristics of L. monocytogenes isolates based on WGS data The complete sequenced genome assembly sizes of the isolates ranged from 2.56-3.13 Mbp with a mean size of 2.97 Mbp (Supplementary Table 2-1). Isolates belonged to one of three different lineages: LI (n=44, serotypes 4b, 1/2b, 3b and 3c), LII (n=121, serotypes 1/2a, 1/2c, and 3a), and LIII (n=1, serotype 4c). The majority of isolates were serotype 1/2a (n=92), followed by 1/2c and 4b (n=25 each), 1/2b (n=18), 3a (n=2), and 3b and 4c (n=1 each) (Table 2-1). The exact serotype was not determined for two remaining isolates. Beyond serotypes, our isolates belonged to 36 different known STs and a further nine were assigned novel STs (ST1017-1025). Isolates also belonged to one of 29 different CCs with a further seven isolates being unique non-clonal singletons. The most prevalent CCs in the collection were CCs 9, 8, and 7 (Table 2-2). Other less common CCs in decreasing prevalence included CCs 11, 155, 1, 3, and 321 (Table 2-2). Interestingly, only one CC121 isolate existed in our collection. This is surprising given that CC121 is often highly prevalent among L. monocytogenes food-associated isolates (Chenal-Francisque et al., 2011; Ebner et al., 2015; Martín et al., 2014; Maury et al., 2016; Parisi et al., 2010).  The plasmid replication gene, repA, was observed in 55% (n=92) of our isolates, with a prevalence of 41 and 61% among LI and LII isolates, respectively. Notably, one isolate was  71  observed to contain two putative plasmids as indicated by the presence of two different repA containing contigs of 61 and 69 kbp. Among serotypes, repA was present in 100% of 3a isolates, 84% of 1/2c isolates, 78% of 1/2b isolates, 53% of 1/2a isolates, and 16% of 4b isolates (Table 2-1). Among CCs, plasmids were observed in >80% of CC 3, 5, 9, 11, and 321 isolates (Table 2-2). A phylogeny, constructed on repA sequences as described in Kuenne et al. (2010) , divided the sequences into two groups. Group 1 included isolates from serotypes 1/2a, 1/2b, 1/2c, and 4b with estimated plasmid sizes ranging from 26-88 kbp. Group 2 included 1/2a, 1/2b, 1/2c, and 3a serotype isolates, harbouring significantly larger plasmids (p<0.0005, 55-100 kbp) than those from group 1. These sizes are in line with those observed in Kuenne et al. (2010), supporting the assertion that these contigs belong to plasmids. The most prevalent plasmid size (56553-56554 bp) was observed for 26 isolates from seven different CCs and from both lineages I and II. Also noteworthy is that isolates from Switzerland and Canada contained plasmids with 100% nucleotide identity. PMSCs in inlA were observed in 20% of our isolates encompassing seven (Supplementary Table 2-1) of 19 published PMSCs (Hu et al., 2007a; Jonquieres et al., 1998; Nightingale et al., 2008, 2005a, Olier et al., 2003, 2002; Rousseaux et al., 2004; Van Stelten et al., 2010; Van Stelten and Nightingale, 2008; Wu et al., 2016) and one novel PMSC at 760aa, which was identified in two serotype 1/2c isolates. The most common PMSC occurred at 9aa (n=13) and was associated with CC9, serotype 1/2c isolates. Ten of the 14 remaining CC9 isolates also had one of four inlA PMSCs (326, 576, 685, 760aa) and all CC321 isolates contained inlA PMSC’s at 700aa (Supplementary Table 2-1). An additional thirteen isolates, all from the serotype 4b CCs 6 and 315, contained a three codon deletion mutation previously reported in Kovacevic et al. (2013).  72  With the exception of CCs 5 and 9, all isolates from the same CC either contained full length inlA or a truncated version. During the screening of the whole genome sequences, the absence of lmo1078 was noted among serotype 4b isolates. This gene, which encodes a UDP-glucose pyrophosphorylase, has been previously demonstrated to have a role in L. monocytogenes cold growth (Chassaing and Auvray, 2007). It was also observed that 70% (n=116) of strains possessed SSI-1 with this island being most prevalent among serotype 1/2c isolates (100%) followed by 1/2b (94%), 1/2a (71%), and 4b (16%) (Table 2-2). Furthermore, all isolates from CCs 3, 5, 7, 8, 9, 155, 224, 315 and 321 contained SSI-1 (Table 2-2). All remaining isolates possessed a homolog to F2365_0481 in place of SSI-1 (Ryan et al., 2010), with the exception of the CC121 isolate which possessed lin0464 and lin0465 homologs as reported in Hein et al. (2011). The LGI1 indicator gene, emrE, was found in 16 of our isolates and as previously reported (Althaus et al., 2014; Gilmour et al., 2010; Kovacevic et al., 2015), all originated from Canada and 14 were serotype 1/2a ST120-CC8. The remaining two isolates represented novel STs (ST1022 and 1025) that also belonged to CC8. All emrE containing isolates also harboured SSI-1 and full length inlA.     73  Table 2-1. Genetic characteristics and prevalence of sensitive and tolerant phenotypes among L. monocytogenes belonging to different serotypes.  Serotype n (%) Plasmid+ (%)a Full length inlA (%)a SSI-1+ (%)a CS (%)a CT (%)a SS (%)a ST (%)a AS (%)a AT (%)a DS (%)a DT (%)a 4b 25 (15) 4 (16) 14 (56) 4 (16) 4 (16) 2 (8) 1 (4) 3 (12) 1 (4) 4 (16) 2 (8) 3 (12) 1/2b 18 (11) 14 (78) 15 (83) 17 (94) 0 2 (11) 3 (17) 2 (11) 0 7 (39) 2 (11) 4 (22) 1/2a 92 (55) 49 (53) 85 (92) 65 (71) 4 (4) 11 (12) 18 (20) 12 (13) 22 (24) 7 (8) 10 (11) 11 (12) 1/2c 25 (15) 21 (84) 3 (12) 25 (100) 5 (20) 3 (12) 4 (16) 0 2 (8) 4 (16) 4 (16) 4 (16) 3a 2 (1) 2 (100) 0 2 (100) 0 0 1 0 1 0 0 0 3b 1 1 1 1 0 0 0 0 0 0 0 1 4c 1 0 0 0 0 0 0 0 0 0 1 0 1/2b, 3b, 7 1 0 1 1 0 0 0 0 0 0 0 0 1/2a, 3a 1 1 0 1 0 0 1 0 0 0 1 0 Sum 166 92 119 116 13 18 27 17 26 22 20 23 CS, cold sensitive; CT, cold tolerant; SS, salt sensitive; ST, salt tolerant; AS, acid sensitive; AT, acid tolerant; DS, desiccation sensitive; DT, desiccation tolerant a Percentages relate to prevalence within the serotype    74  Table 2-2. Genetic characteristics and prevalence of sensitive and tolerant phenotypes among L. monocytogenes belonging to different clonal complexes.  CC Associated Serotypes n (%) Plasmid+ (%)a Full length inlA (%)a SSI-1+ (%)a CS (%)a CT (%)a SS (%)a ST (%)a AS (%)a AT (%)a DS (%)a DT (%)a 1 4b 6 (4) 0 6 (100) 0 2 (33) 0 1 (17) 0 0 1 (17) 1 (17) 0 2 4b 3 (2) 0 3 (100) 0 1 (33) 0 0 2 (67) 0 0 0 0 3 1/2b 6 (4) 5 (83) 6 (100) 6 (100) 0 0 1 (17) 0 0 1 (17) 1 (17) 1 (17) 4 4b 3 (2) 0 3 (100) 0 0 1 (33) 0 0 0 2 (67) 1 (33) 0 5 1/2b 7 (4) 7 (100) 4 (57) 7 (100) 0 1 (14) 0 1 (14) 0 4 (57) 1 (14) 1 (14) 6 4b 7 (4) 4 (57) 0 0 1 (14) 0 0 0 1 (14) 0 0 2 (29) 7 1/2a 17 (10) 10 (59) 17 (100) 17 (100) 2 (12) 1 (6) 8 (47) 0 5 (29) 0 1 (6) 3 (18) 8 1/2a 22 (13) 14 (64) 22 (100) 22 (100) 0 3 (14) 2 (9) 1 (5) 1 (5) 3 (14) 2 (9) 0 9 1/2c, 1/2a 27 (16) 23 (85) 4 (15) 27 (100) 5 (19) 4 (15) 6 (22) 0 3 (11) 4 (15) 6 (22) 4 (15) 11 1/2a 11 (7) 10 (91) 11 (100) 0 0 1 (9) 0 4 (36) 0 1 (9) 0 4 (36) 14 1/2a 3 (2) 0 3 (100) 0 0 1 (33) 0 1 (33) 0 0 0 0 20 1/2a 3 (2) 0 3 (100) 0 0 1 (33) 1 (33) 1 (33) 2 (67) 0 0 0 155 1/2a 8 (5) 1 (13) 8 (100) 8 (100) 1 (13) 0 2 (25) 1 (13) 3 (38) 1 (13) 1 (13) 0 224 1/2b 4 (2) 0 4 (100) 4 (100) 0 1 (25) 2 (50) 0 0 1 (25) 0 3 (75) 315 4b 4 (2) 0 0 4 (100) 0 0 0 0 0 0 0 1 (25) 321 1/2a, 3a 6 (4) 6 (100) 0 6 (100) 0 0 0 1 (17) 5 (83) 0 0 2 (33) Other 1/2a, 1/2b, 4b, 4c 29 (17) 12 (41) 25 (86) 15 (52) 1 (3) 4 (14) 4 (14) 5 (17) 6 (20) 4 (14) 6 (20) 2 (7) Sum  166 92 119 116 13 18 27 17 26 22 20 23 CS, cold sensitive; CT, cold tolerant; SS, salt sensitive; ST, salt tolerant; AS, acid sensitive; AT, acid tolerant; DS, desiccation sensitive; DT, desiccation tolerant a Percentages relate to prevalence within CCs        75  2.3.2 Stress tolerance distributions among L. monocytogenes isolates All L. monocytogenes isolates were evaluated on their ability to grow in cold (4°C), salt (6% NaCl), and acid (pH 5) stress conditions as well as survive desiccation stress (33% RH). The cold growth plate count data was modelled using the Dalgaard and Koutsoumanis (2001) logistic model because it was more accommodating of fewer sampling points [average R2=0.998, mean standard error (MSE) =0.129]. From the std-µmax values, 13 isolates were classified as cold sensitive and 18 were classified as cold tolerant with average std-µmax values of 0.85±0.08 and 1.09±0.02, respectively (Fig 2-1A). The Baranyi and Roberts model (1994) suitably modelled the data from the salt, acid, and desiccation tolerance assays,  with average R2 and MSE values ranging from 0.997-0.998 and 0.003-0.017, respectively. Overall, 27 and 17 isolates were classified as salt sensitive and tolerant, respectively, with average std-µmax values of 0.83±0.05 and 1.16±0.05 (Fig 2-1B); 26 and 22 isolates were classified as acid sensitive and tolerant with average std-µmax values of 0.64±0.14 and 1.34±0.12 (Fig 2-1C); and 20 and 23 isolates were identified as desiccation sensitive and tolerant isolates with average std-TRGs of 0.81±0.06 and 1.22±0.11 (Fig 2-1D), respectively.  76   Figure 2-1. Stress tolerance distributions of 166 L. monocytogenes isolates. std-µmax of isolates grown in (A) BHIB at 4°C, (B) BHIB+6% NaCl at 25°C, and (C) BHIB pH 5 at 25°C. (D) std-TRG of isolates after being desiccated at 33% RH for 3 days in BPW at 20°C and then rehydrated with BHIB and grown at 30°C. Isolates were classified as sensitive or tolerant if they displayed an average std-µmax or std-TRG >1 SD from the median (=1). std-µmax, standardized maximum growth rate; std-TRG, standardized time to detectable regrowth; BHIB, brain heart infusion broth; BPW, buffered peptone water.   2.3.3 Overlapping stress-tolerance phenotypes Five isolates were classified as sensitive to three out of four stresses and 10 isolates were sensitive to two stresses, seven of which were salt and acid sensitive (Fig 2-2A). Only two isolates were classified as tolerant to three out of the four stresses and another 14 isolates were tolerant to two of the four stresses (Fig 2-2B). Twenty additional isolates displayed a total of 16 combinations  77  of overlapping sensitive and tolerant phenotypes (Supplementary Table 2-1). The most common overlapping phenotypes were salt and acid sensitive (n=10), salt sensitive and desiccation tolerant (n=6), cold and acid tolerant (n=5), and cold tolerant and salt sensitive (n=5).   Isolates designated sensitive, intermediate or tolerant to one stress were analyzed to determine if they significantly differed in their tolerances to other stresses. When grown in 6% NaCl, acid tolerant isolates had larger std-µmax values compared to acid sensitive isolates (p=0.03, x̄=1.02 vs. 0.95). Similarly, salt sensitive isolates had smaller std-µmax values (x̄=0.80) in BHIB pH 5 compared to intermediate (p<0.0005, x̄=1.03) and salt tolerant isolates (p=0.006, x̄=1.00).   Figure 2-2. Numbers of L. monocytogenes isolates with multiple sensitivities or tolerances to food-related stresses. (A) Sensitive isolates. (B) Tolerant isolates.   2.3.4 Stress tolerances of L. monocytogenes lineages, serotypes, and clonal complexes  Between lineages, the only significant difference observed was that LI isolates had significantly larger std-µmax values in BHIB pH 5 than LII isolates (p<0.0005, x̄=1.13 vs. x̄=0.94). Additional significant differences were observed between serotypes. At 4°C, serotype 1/2a isolates had significantly larger (p=0.017) std-µmax values compared to serotype 1/2c isolates (Fig 2-3). In  78  support of this, serotype 1/2a isolates accounted for 61% of the cold tolerant isolates and only 31% of cold sensitive isolates compared to a 55% prevalence of this serotype in the collection (Table 2-1). Similarly, serotype 1/2c isolates accounted for 38% of cold sensitive isolates despite a 15% overall prevalence in the collection (Table 2-1). When isolates were grown in 6% NaCl, no significant differences were observed between serotypes (Fig 2-3); however, 71% of salt tolerant and 67% of salt sensitive isolates were serotype 1/2a isolates, relative again to a prevalence of 55% in the collection (Table 2-1).  In BHIB pH 5, serotype 1/2a isolates had significantly smaller (p≤0.027) std-µmax values than serotypes 1/2b, 1/2c, and 4b (Fig 2-3). In agreement with these findings, 85% of acid sensitive isolates were serotype 1/2a whereas only 32% of acid tolerant isolates were serotype 1/2a (Table 2-1). No significant differences (p>0.05) were observed between serotypes with respect to desiccation stress std-TRGs (Fig 2-3). Beyond the stress tolerances of lineages and serotypes, some significant differences were also detected between CCs. As a minimum of six isolates per CC were needed to ensure a power >0.80 for ANOVA results, statistical analyses were only performed using CCs 1, 3, 5, 6, 7, 8, 9, 11, 155, and 321. Figure 4 shows the average levels of cold (std-µmax), salt (std-µmax), acid (std-µmax), and desiccation (std-TRG) tolerance among CCs with three or more isolates. At 4°C, no significant differences were found between the growth rates of different CCs; however, it was interestingly to see that CCs associated with 4b isolates had both the lowest and highest average std-µmax values at 4°C, demonstrating why stress tolerance differences were not observed between this serotype and others at 4°C (Fig 2-4A).  In 6% NaCl, CC7 (1/2a) isolates had significantly (p<0.05) smaller std-µmax values compared to CCs 5 (1/2b), 8 (1/2a), 11 (1/2a), and 155 (1/2a) (Fig 2-4B). This highlights the range  79  of salt tolerances between CCs within the same serotype and again explains why no significant differences were observed at the serotype level for salt tolerance. In support of the results shown in Fig 2-4B, 67% of CC2 isolates were salt tolerant while 50% of CC224, 47% of CC7, and 22% of CC9 isolates were salt sensitive (Table 2-2).  In BHIB pH 5, CC5 (1/2b) isolates exhibited significantly (p<0.05) larger std-µmax values than CCs 7, 155, and 321, and CC321 isolates additionally had smaller (p<0.05) std-µmax values compared to CCs 1, 3, and 11 (Fig 4C). CC1 (4b) isolates also had significantly (p=0.02) smaller std-µmax values compared to CC7 (1/2a) isolates. From Fig 2-4C it can be seen that lineage I isolates were more acid tolerant than LII isolates, as the five CCs with the highest average std-µmax values in BHIB pH 5 were all from LI while the five CCs with the lowest average std-µmax values belonged to LII (predominantly 1/2a isolates). Notably, 57% of CC5 and 67% of CC4 isolates were acid tolerant while 83% of CC321, 67% of CC20, and 29% of CC7 isolates were acid sensitive (Table 2-2).  No significant differences were found between the desiccation stress std-TRGs of different CCs (Fig 2-4D). Nevertheless, CC224 (1/2b) had the smallest average std-TRGs and correspondingly, 75% of these isolates were classified as desiccation tolerant. CC11 (1/2a) had the next smallest std-TRGs while CCs 1 and 4 (both 4b) had the two largest average std-TRGs.  80   Figure 2-3. Levels of tolerance to food-related stresses among L. monocytogenes serotypes. Isolates were evaluated on their ability to survive cold (BHIB at 4°C), salt (BHIB+6% NaCl, 25°C), acid (BHIB pH 5, 25°C), and desiccation stress (33% RH for 3 days at 20°C followed by rehydration with BHIB at 30°C). Error bars represent standard deviations (n-1). Serotypes with different letters within the same stress are significantly different (p<0.05). std-µmax, standardized maximum growth rate; std-TRG, standardized time to detectable regrowth; BHIB, brain heart infusion broth.  81   Figure 2-4. Levels of tolerance to food-related stresses of different L. monocytogenes clonal complexes. Isolates were evaluated on their ability to survive (A) cold (BHIB at 4°C), (B) salt (BHIB+6% NaCl, 25°C), (C) acid (BHIB pH 5, 25°C), and (D) desiccation stress (33% RH for 3 days at 20°C followed by rehydration with BHIB at 30°C). Error bars represent standard deviations (n − 1) of standardized model values. CCs with different cases of the same letter are significantly different (p < 0.05). std-μmax, standardized maximum growth rate; std-TRG, standardized time to detectable regrowth; BHIB, brain heart infusion broth. 2.3.5 Associations between plasmid harbourage and stress tolerances Although plasmids were identified in 55% of all isolates, a higher percentage of plasmid carriers were observed among acid tolerant (73%), desiccation sensitive (75%), and desiccation tolerant (60%) isolates as compared to cold tolerant (33%) and acid sensitive (46%) isolates (Fig 2-5). Within LII, plasmid-positive isolates had smaller std-µmax values at 4°C (p=0.024, x̄= 1.00 vs. 1.02) and larger std-µmax (p<0.0005, x̄=1.01 vs. 0.86) values when grown in BHIB pH 5 compared to their plasmid-free counterparts. No significant differences were found between the  82  stress tolerance levels of LI plasmid-harbouring and plasmid-free isolates. However, it was observed that isolates containing repA group 1 plasmids; which were significantly (p<0.0005) smaller than group 2 plasmids, had smaller std-µmax (p=0.002, x̄=0.98 vs. 1.04) values in 6% NaCl.  Figure 2-5. Prevalence of full length inlA and plasmid harbourage among L. monocytogenes stress-tolerance phenotypes. (A) Cold sensitive (CS), intermediate (CI), and tolerant (CT) isolates. (B) Salt sensitive (SS), intermediate (SI), and tolerant (ST) isolates. (C) Acid sensitive (AS), intermediate (AI), and tolerant isolates (AT). (D) Desiccation sensitive (DS), intermediate (DI), and tolerant (DT) isolates. Full length inlA and the presence of a plasmid were observed in 72% and 55% of all isolates, respectively.   2.3.6 Associations between inlA profiles and stress tolerances Full length inlA was observed in 72% of isolates, where a higher percentage of the intact gene prevailed among cold (89%), salt (94%) and acid (82%) tolerant isolates while a lower prevalence was detected among desiccation tolerant isolates (35%) (Fig 2-5). Statistically, isolates  83  with full length inlA had significantly larger std-µmax values at 4°C than isolates with an inlA PMSC (p=0.001, x̄=1.01 vs. 0.97). Additionally, serotype 4b isolates possessing a three-codon deletion in inlA, had significantly shorter desiccation stress std-TRGs (p=0.002 x̄=0.94 vs. 1.05) compared to serotype 4b isolates with full length inlA. No significant associations were found between inlA profiles and salt or acid stress tolerance.   2.3.7 Associations between stress tolerances and the presence of SSI1 or LGI1  In 6% NaCl, isolates containing SSI-1 had significantly smaller std-µmax values (p=0.004, x̄=0.98 vs. 1.03) than isolates without SSI-1 though this difference was not large. To determine potential associations between LGI1 harbourage and stress tolerance, serotype 1/2a isolates containing the island were compared to other 1/2a LGI1-negative isolates but similar to SSI-1, no significant differences in stress tolerances were detected.  2.3.8 SNP analyses of stress-sensitive and tolerant isolates Fig 2-6 shows a whole genome SNP phylogeny of all 166 L. monocytogenes isolates with their corresponding genetic and phenotypic properties. In this figure, groups of closely related isolates that share the same phenotypes can be seen. However, also shown are several cases where neighbouring isolates have opposing stress tolerances. Of particular interest was whether specific SNPs could be related to isolates possessing the same stress-tolerance phenotypes; however, none were detected to be uniquely shared among stress tolerant isolates that weren’t also seen in intermediate or sensitive isolates. Among stress sensitive isolates, unique SNPs shared by subsets of isolates were identified, but no single SNP was prevalent among >4 isolates from the same stress sensitive phenotype group. In contrast, a large number of SNPs were uniquely observed for one or two isolates from the same stress sensitive group, causing frameshifts, premature stop  84  codons, loss of start codons, or missense variants. Information regarding the SNPs identified among all sensitive and tolerant isolates are presented in Supplentary 2-2 to 2-9. Notably, a number of stress sensitive isolates contained different PMSCs in several σB regulator genes. A cold and desiccation sensitive isolate contained a PMSC in rsbS as did two other desiccation sensitive isolates. Furthermore, an additional cold sensitive isolate contained a PMSC in rsbV, and two desiccation sensitive isolates contained PMSCs in rsbU.  85   Figure 2-6. Whole genome single nucleotide polymorphism (SNP) phylogeny of 166 L. monocytogenes isolates and their associated genetic characteristics and stress-tolerance phenotypes. The scale at the bottom indicates the substitutions per SNP.  86  2.3.9 Genomic islands of stress-sensitive and tolerant isolates All L. monocytogenes isolates were predicted to harbour 1,318 genomic islands in total, resulting in an average of eight genomic islands per genome. These islands were clustered into 200 groups of similar sequences. The conservation of genomic island groups across L. monocytogenes lineages ranged from unique to a single isolate to conserved in 97 isolates (58%). Most frequently, intermediate groups conserved in subsets of monophyletic isolates were observed, as can be expected from a combination of vertical inheritance of the genomic islands with further modifications by mutation, insertions and deletions. The clustering of L. monocytogenes isolates based on the presence or absence of groups of genomic islands reflected their phylogenetic proximity and did not relate particular genomic island content in isolates with the exhibition of similar phenotypes. Indeed, no single genomic island was found to occur in a large proportion of strains with a given phenotype (Fig A-1, Appendix A).     2.4 Discussion 2.4.1 The tolerance of L. monocytogenes to food-related stresses differs between and within lineages, serotypes, and clonal complexes 2.4.1.1 Cold stress The ability of L. monocytogenes to grow at refrigeration temperatures highlights this pathogen as a concern for the food industry and consumers alike. However, it is known that L. monocytogenes strains can largely differ in their ability to adapt to cold stress. In the present study, it was found that serotypes 1/2a and 1/2b were on average more cold-tolerant than serotypes 4b and 1/2c. Other cold growth studies have also reported serotype 1/2a strains to be more cold tolerant than serotype 4b strains (Bunčić et al., 2001; Junttila et al., 1988; Lianou et al., 2006)  87  though similar to the current findings, many of these differences were not statistically significant due to strain to strain variations. Barbosa et al. (1994) reported that out of 39 L. monocytogenes strains, Scott A, a 4b clinical isolate, grew the slowest at 4°C and that 1/2a strains grew the fastest followed by 1/2b, and 4b. Similarly, De Jesús & Whiting (2003) reported that LII isolates (all serotype 1/2a) exhibited the shortest LPDs at 5°C followed by LI and then LIII isolates. Researchers have suggested that LII strains may be able to survive better under food-related stresses due to an enhanced ability to acquire advantageous mutations and extrachromosomal DNA compared to LI strains which typically have more conserved genomes (Dunn et al., 2009; Orsi et al., 2011, 2008a, 2007; Ragon et al., 2008). Certain stress response genes, predominantly involved in membrane transport and cell wall structure (Doumith et al., 2004), have also been reported to be present in LII isolates but absent among LI isolates (Borucki and Call, 2003; Call et al., 2003; Chan and Wiedmann, 2008; Doumith et al., 2004; Zhang et al., 2003). Given the critical roles of these structures in allowing bacteria to adapt and tolerate numerous stresses (Álvarez-Ordóñez et al., 2008; Annous et al., 1997; Klein et al., 1999; Verheul et al., 1997; Weber et al., 2001), it is not surprising that L. monocytogenes lineages and serotypes can behave differently under certain stresses. The alternative sigma factor, σC, and lmo1078, encoding a UDP-glucose pyrophosphorylase, are examples of genes with reported roles in L. monocytogenes cold tolerance, that are present in LII strains but absent in LI and serotype 4b strains, respectively (Chan and Wiedmann, 2008; Chassaing and Auvray, 2007). These absences may partly explain the overall reduced cold tolerance of serotype 4b isolates respective to 1/2a isolates. However, they do not appear to be necessary for adequate cold growth as in the present study some 4b isolates were also classified as cold tolerant.   88  2.4.1.2 Salt stress LI strains have been shown to be more salt tolerant than LII strains (Bergholz et al., 2010) and serotype 4b strains to be more salt tolerant than serotype 1/2a and 1/2b strains (Bergholz et al., 2010; Ribeiro et al., 2014; van der Veen et al., 2008). In the present study, no significant differences (p>0.05) were found between the growth rates of different serotypes in 6% NaCl; however, five times as many 4b isolates were classified as salt tolerant as were classified as salt sensitive. Additionally, despite a 55% prevalence of serotype 1/2a isolates in our collection, isolates of this serotype accounted for 71% of salt tolerant isolates but also for 67% of salt sensitive isolates, indicating a broad range of salt tolerance among isolates from this serotype. These differences were found to be associated with specific 1/2a CCs, notably, CC7 isolates were on average the most salt sensitive CC and significantly differed from the 1/2a CCs 8 and 11 with most 1/2a salt tolerant isolates belonging to CC11.      2.4.1.3 Acid stress LI isolates tolerated acid stress conditions significantly better than LII isolates. Specifically, serotype 1/2a showed higher sensitivity to acidity. This trend was also clear when the acid tolerance of individual CCs was investigated. van der Veen et al. (2008), also reported that 4b (LI) isolates had enhanced acid tolerance relative to 1/2a isolates. The authors hypothesized that the increased survival of 4b strains may in part be due to the presence of ORF2110 which encodes a putative serine protease similar to HtrA. This protein has been shown to be important for growth at low pH, high osmolarity, and high temperatures (Wonderling et al., 2004; Stack et al., 2005; Wilson et al., 2006). Though this gene may contribute to the overall acid tolerance of 4b isolates, some acid sensitive 4b isolates were also identified. Similarly, despite serotype 1/2a  89  isolates having relatively low acid tolerance overall, some 1/2a isolates were also acid tolerant, highlighting the importance of not overgeneralizing isolate phenotypes based on the trends seen for their sero- or sequence type. 2.4.1.4 Desiccation stress Desiccation tolerance can be described as a bacteria’s ability to survive on a surface for extended periods of time with little access to nutrients and water. Accordingly, persistent strains of L. monocytogenes continuously isolated from food production plants are hypothesized to possess enhanced desiccation tolerance (Vogel et al., 2010). To date, surprisingly little research has been conducted regarding the desiccation survival of L. monocytogenes and that which does exist has focused primarily on factors influencing the survival of a small number of isolates (Truelstrup Hansen and Vogel, 2011; Hingston et al., 2013; Overney et al., 2016; Takahashi et al., 2011; Vogel et al., 2010). The current study is the first to our knowledge, to compare the desiccation tolerance of L. monocytogenes isolates from multiple serotypes. No significant differences (p>0.05) were found between serotypes or CCs; however, some prominent trends were observed. Serotypes 1/2c and 1/2b were on average the most desiccation tolerant, followed by 4b and 1/2a. More specifically, CC224 (1/2b) isolates had the highest levels of desiccation survival. Interestingly, a large listeriosis outbreak in Denmark, which resulted in 41 illnesses and 17 deaths, was linked to the consumption of deli meat contaminated with a CC224 strain (Kvistholm Jensen et al., 2016). Though there is not enough evidence to suggest that all CC224 strains are desiccation tolerant, it is possible that long-term desiccation survival may have contributed to the occurrence of this outbreak. Another interesting finding from the present study was that the most desiccation sensitive isolate, deviating more than 4.5 SD from the median, was a CC193 (serotype 1/2a)  90  isolate. Since this was the only CC193 isolate in our collection, it would be interesting to analyze additional isolates from this CC to determine if this sequence type is associated with a high degree of desiccation sensitivity. 2.4.2 Certain genetic elements are associated with the stress tolerance of L. monocytogenes 2.4.2.1 Plasmids The presumptive presence of a plasmid(s) was detected in 55% of our isolates which is comparable to other studies where rates of plasmid isolation have ranged from 0-79% with an overall average of around 30% (Kolstad et al., 1992; Lebrun et al., 1992; McLauchlin et al., 1997; Perez-Diaz et al., 1982; Peterkin et al., 1992). In agreement with earlier work, we also observed that plasmid DNA was more prevalent among LII isolates than LI isolates (Kolstad et al., 1992; Lebrun et al., 1992; Margolles et al., 1998; McLauchlin et al., 1997; Orsi et al., 2011). Kuenne et al. (2010) discovered that L. monocytogenes plasmids could be divided into two phylogenetic groups based on their repA sequences and that plasmids belonging to the second group were larger (77-83 kbp) than those belonging to the first group (32-57 kbp). In the present study, plasmid repA sequences also formed two distinct phylogenetic groups and in agreement with Kuenne et al. (2010), group 2 plasmids were significantly (p<0.05) larger (55-100 kbp) than group 1 plasmids (26-88 kbp). Notably, one serotype 1/2b isolate contained two plasmids of similar sizes (62 and 69 kbp) but belonging to different repA groups. Though rare, the presence of two plasmids has been reported in other Listeria spp. isolates (Earnshaw and Lawrence, 1998; Margolles et al., 1998).  Our results showed that among LII isolates, which exhibited higher rates of plasmid harbourage than LI isolates, plasmid harbourage was associated with significantly enhanced acid  91  tolerance but also cold sensitivity. Studies have shown that plasmid harbourage and subsequent replication increases the metabolic demands of cells, leading to decreased growth rate relative to plasmid-free strains [reviewed in Diaz Ricci and Hernández (2000)]. However, depending on the genes contained on a plasmid, plasmid harbourage can also provide cells with a growth advantage when exposed to certain conditions. In this study, isolates with the larger repA group 2 plasmids were significantly more salt-tolerant than isolates that harboured the smaller repA group 1 plasmids, collectively suggesting that plasmid harbourage may be a hindrance to L. monocytogenes during replication at low temperatures but provide an advantage when exposed to acid and salt stress conditions. Furthermore, the observation that isolates containing larger plasmids had higher levels of salt tolerance suggests that these plasmids may contain additional genes that are beneficial for adaptation to high osmolarity environments.  To date, plasmids acquired by L. monocytogenes have been shown to contain genes that confer resistance to benzalkonium chloride (bcrABC) (Elhanafi et al., 2010; Katharios-Lanwermeyer et al., 2012; Rakic-Martinez et al., 2011), cadmium (cadA2, cadAC) (Katharios-Lanwermeyer et al., 2012; Lebrun et al., 1992; Rakic-Martinez et al., 2011) and antibiotics including chloramphenicol, clindamycin, erythromycin, streptomycin and tetracycline (Hadorn et al., 1993; Poyart-Salmeron et al., 1990). Listeria spp. plasmids also commonly contain several other uncharacterized efflux pumps (MDR, SMR, MATE) (Boylan et al., 2006; Gibson et al., 2000; Kuroda and Tsuchiya, 2009; Masaoka et al., 2000), as well as oxidative stress response genes (peroxidases, reductases) (Kuenne et al., 2010) but their exact roles in stress tolerance have yet to be investigated. In other bacterial species, multidrug efflux pumps have been linked to stress response, virulence, and quorum sensing [reviewed in Li and Nikaido (2009)]. Ma et al. (1995) reported that transcription of a MDR pump (acrAB) in E. coli increased in response to fatty acids,  92  ethanol, high salt, and cellular transitioning into stationary phase. Among the putative plasmid associated contigs a wide variety of different genes were identified including those encoding cell surface proteins, lipoproteins, secretion pathways, heavy metal transporters, transcription regulators, general stress proteins (CplB, ClpL), NADH oxidoreductases, a glycine betaine transport permease (ProW), and the multidrug resistance proteins EbrA and EbrB among others. All group 2 plasmids shared a general secretion pathway protein and a cell surface protein. Other genes identified among many but not all group 2 plasmids included those which encoded DNA topoisomerase III, a membrane-bound protease, a NLP/P60 family lipoprotein, an NADH peroxidase, and a type IV secretory pathway. Further investigations are currently focusing on whether these genes or others with no known function contribute to the acid and salt tolerance of L. monocytogenes.  Lastly, it should be highlighted that plasmids with 99-100% nucleotide identity were found in isolates from different serotypes, provinces (Alberta and British Columbia), and countries (Canada and Switzerland). One plasmid was observed in 26 isolates, which strongly suggests that L. monocytogenes benefits from its presence. The occurrence of the same plasmid in multiple food-related isolates from different regions also suggests that bacteria are frequently transported between places of food production, possibly alongside imported raw materials. In constrast, other plasmids were also found to be conserved among strains from specific CCs, serotypes, provinces and countries.   93  2.4.2.2 Full length inlA Full length inlA profiles were observed among 92% of serotype 1/2a isolates, 83% of serotype 1/2b isolates, and 12% of serotype 1/2c isolates, reflecting what has been previously observed (Felicio et al., 2007; Jacquet et al., 2004; Jonquieres et al., 1998; Nightingale et al., 2005a; Orsi et al., 2007; Ragon et al., 2008; Rousseaux et al., 2004). Additionally, 44% of 4b isolates contained a 3-codon deletion which unlike inlA PMSCs, is not associated with attenuated virulence (Kanki et al., 2015; Kovacevic et al., 2013).  The present study results showed that full length inlA profiles were more prevalent among cold, salt, and acid tolerant isolates compared to their sensitive counterparts. Also, isolates with full length inlA profiles were significantly (p<0.05) more cold tolerant than isolates containing inlA PMSCs. Kovacevic et al. (2013) were the first to report that cold tolerant isolates are more  likely to possess full length inlA than intermediate and cold sensitive isolates. This increased stress tolerance has now been shown to extend to salt and acid tolerance, making it reasonable to hypothesize that full length inlA may participate in the L. monocytogenes stress response. When bacteria are exposed to unfavorable conditions, their cell envelope is the first line of defense. It is possible that the absence of cell wall anchored InlA proteins may alter cell-surface characteristics, leaving cells more susceptible to certain environmental stresses. Interestingly, only a small percentage of desiccation tolerant isolates contained full length inlA while serotype 4b isolates with full length inlA profiles had significantly (p<0.05) impaired desiccation survival relative to those with a 3-codon deletion. Again, it is suspected that the structure of inlA may influence the desiccation tolerance of L. monocytogenes, this time with the full-length form possibly imparting a disadvantage. Other researchers have also detected associations between internalin mutations and certain phenotypes. In Hingston et al. (2015), inlC was identified as the interrupted gene in a  94  desiccation tolerant transposon mutant and Franciosa et al. (2009) found that strains possessing a truncated InlA formed increased levels of biofilm. Similarly, transposon mutants containing an interrupted internalin A, B, or H gene, formed thicker biofilms relative to the wildtype (Piercey et al., 2016). Together, these findings along with those presented in this study, emphasize a need for more research regarding the potential roles of internalins in the responses of L. monocytogenes to certain stresses.   2.4.2.3 SSI-1  SSI-1 is a five-gene cluster which has previously been shown via mutagenesis studies to enhance L. monocytogenes tolerance to acid, salt, and low temperature conditions (Cotter et al., 2005; Ryan et al., 2010). On the other hand, Arguedas-Villa et al. (2014) found no significant differences in cold tolerance between naturally occurring L. monocytogenes isolates with and without SSI-1. In the present study, it was found that SSI-1-positive isolates showed no enhanced cold, acid, salt and desiccation stress tolerances relative to SSI-1 negative isolates. It is possible that any positive influence of SSI-1 on the stress tolerance of L. monocytogenes may be masked by the presence of other genetic elements when comparing large collections of isolates as opposed to a mutant and its wildtype strain.  2.4.2.4 LGI1 LGI1 is a Listeria 50 kb genomic island which was first identified in Canadian CC8 isolates associated with a large 2008 listeriosis outbreak involving contaminated deli meats and resulting in 22 fatalities (Gilmour et al., 2010). Since then, LGI1 has been identified in other CC8 L. monocytogenes isolates from Canada (Kovacevic et al., 2013) but not from other countries (Althaus et al., 2014). In agreement with these studies, the presence of LGI1 was only detected in  95  Canadian isolates from CC8. However, instead of all LGI1+ isolates being ST120 as previously reported, two novel CC8 STs (ST1022 and 1025) were also associated with LGI1 harbourage. The conservation of LGI1 among Canadian isolates and its association with a fatal outbreak has led to heightened interest in the putative functions of the genes located on this island including those encoding putative type II and type IV secretion systems, pilus-like surface structures, a multidrug efflux pump homologue (EmrE), and an alternative sigma factor (Gilmour et al., 2010). Recently, Kovacevic et al. (2015) reported that deletion of LGI1 genes with putative efflux (emrE), regulatory (lmo1852), and adhesion (sel1) functions, had no impact on the tolerance of L. monocytogenes to acid, cold, or salt, but that deletion of emrE increased susceptibility to quaternary ammonium-based sanitizers. Based on these findings, it was investigated whether the presence or absence of the whole LGI1 island could be associated with stress tolerance differences between CC8 isolates; however, no significant differences (p>0.05) were identified. Consequently, LGI1 had no major influence on the ability of L. monocytogenes to adapt to the food-related stresses evaluated in the present study. However, it is possible that the island contributes in other ways to the persistence of CC8 Canadian isolates in food-processing environments in addition to the role of emrE in sanitizer resistance. 2.4.2.5 SNPs associated with stress-tolerance phenotypes An important finding from this study was that closely related isolates from within the same clonal complexes exhibited opposing stress tolerances, suggesting that minor genetic differences can also exert great impact on stress-tolerance phenotypes. This was observed in a study by Hoffmann et al. (2013), where a single thymine deletion in the σA-like promoter region of betL, encoding an osmolyte transporter specific for betaine uptake, dramatically increased betL  96  transcription, and hence the osmo- and chill-tolerance. Karatzas et al. (2003) reported that a spontaneous high hydrostatic pressure tolerant L. monocytogenes mutant of Scott A, contained a single codon deletion in ctsR, a negative regulator of several heat shock and general stress proteins, that also conferred increased thermo-tolerance and resistance to H2O2. Additionally, in Metselaar et al. (2015) a number of spontaneous acid tolerant mutants were found to contain SNPs in the ribosomal protein gene rpsU. None of these aforementioned mutations were detected in the present study.   Analyses to determine if unique SNPs could be detected among isolates from individual stress-tolerance phenotype groups were also performed. Studies identifying the genetic basis of phenotypic traits using the variation within natural populations are known as a genome-wide association studies (GWAS). While GWAS have been effective for identifying mutations responsible for phenotypic traits in humans, the clonal nature of bacterial replication where mutations can reach a high frequency on a single genetic background, makes it difficult to distinguish mutations responsible for an observed phenotype (Falush, 2016; Read and Massey, 2014). As a result, bacterial molecular epidemiology has focused on identifying clonal lineages with particular phenotypic properties rather than identifying the specific genetic variants responsible. Recently, Earle and colleagues (2016) used a GWAS approach to successfully identify genes and genetic variants underlying resistance to 17 antimicrobials in over 3000 isolates of taxonomically diverse clonal and recombining bacteria. While these results show the potential of bacterial GWAS, antimicrobial resistance is usually gained during antimicrobial exposure and thus it is more likely that the traits evolve on multiple independent backgrounds making them easier to detect (Falush, 2016). To date, the identification of mutations responsible for more complex phenotypes such as those evaluated in the present study, remain challenging.  97  Our analyses did not detect any unique SNPs among more than one isolate from the same stress tolerant group, suggesting homoplasy among stress tolerant phenotypes where mutations evolve independently to confer tolerance. On the contrary, a few different SNPs were identified among four or fewer isolates from the same stress sensitive groups. Notably, the cold sensitive phenotypes of two isolates may be associated with PMSCs detected in the σB regulator genes rsbS and rsbV (Voelker et al., 1995). In support of this hypothesis, deletion of rsbV in a L. monocytogenes mutagenesis study resulted in impaired cold stress tolerance (Chan et al., 2008). Three desiccation sensitive isolates also contained different PMSCs in rsbS, including one isolate which was also cold sensitive. An additional two desiccation sensitive isolates contained different PMSCs in another σB posttranscriptional regulator, rsbU. Given the importance of σB in the adaptation of L. monocytogenes to several environmental stresses (Kazmierczak et al., 2003; Wiedmann et al., 1998), it is possible that these mutations contributed to the reduced desiccation tolerance of these isolates. In fact, in Huang et al. (2015) an L. monocytogenes sigB mutant demonstrated reduced desiccation survival relative to the wildtype strain.  2.4.2.6 Genomic islands In this study, specific genomic islands could not be exclusively associated with a particular stress-tolerance phenotype. This is to be expected as genomic islands often contain virulence factors, and in general these are overrepresented in genomic islands as compared to the chromosome (Sui et al., 2009). In L. monocytogenes in particular, genomic islands have been associated with virulence, heavy metal resistance, and benzalkonium chloride efflux (Gilmour et al., 2010; Kovacevic et al., 2015; Kuenne et al., 2010). SSI-1 as described above, has been  98  previously associated with the L. monocytogenes stress response, but was not significantly correlated with the stress-tolerance phenotypes examined in the present study.  2.5 Conclusions In summary, the tolerances of L. monocytogenes to certain food-related stresses differs between serotypes as well as CCs with the latter being a better predictor of isolate salt and acid tolerance but not of cold and desiccation tolerance. To the best of our knowledge, this is the first study to evaluate the stress tolerance of different L. monocytogenes CCs. Other noteworthy findings include potential relationships between the presence of full length inlA and enhanced cold tolerance and the presence of a plasmid and enhanced acid tolerance. On the contrary, the presence of genomic islands including SSI-1 and LGI1, provided isolates with no noticeable advantages under the stresses evaluated in this study. Additional research is needed to confirm the potential roles of full length inlA and plasmid associated genes in the response of L. monocytogenes to various stresses. A whole genome SNP phylogeny of isolate assemblies identified a number of unique SNPs shared by up to four stress sensitive isolates while no common SNPs were observed among stress tolerant isolates. More specifically, six isolates with sensitivity to cold and/or desiccation stress contained PMSCs in σB regulator genes (rsbS, rsbU, rsbV) which may be contributing to these phenotypes.  A number of novel genetic elements were also elucidated in this study including nine new L. monocytogenes STs, a new inlA PMSC, the absence of a cold stress associated gene (lmo1078) in 4b isolates, and several connections between L. monocytogenes CCs and the presence/absence or variations of specific genetic elements. For example, SSI-1 was detected in 100% of isolates  99  from specific CCs, certain plasmid groups and sizes were conserved among isolates from the same CCs, and plasmids with 100% identity were found in isolates belonging to the same CCs but from very different geographical areas. While our isolate collection represented a number of L. monocytogenes CCs previously identified as common among food-related isolates, other CCs were less prevalent or absent from our study of Canadian and Swiss isolates. This highlights the regional prevalence of certain L. monocytogenes genotypes and emphasizes the need for more international collaborative studies.  Collectively, the results suggest that using whole genome sequencing to 1) determine the STs of L. monocytogenes food-related isolates and to 2) screen for the presence of genetic elements such as full length inlA and a plasmid(s), could help food processors and food agency investigators to quickly identify if isolates are likely to possess enhanced tolerances to certain stresses that may be facilitating their long-term survival/persistence in a food-processing environment. The US FDA and CDC are rapidly making whole genome sequencing of foodborne bacterial pathogens a routine part of screening to help link illnesses to contaminated foods and to identify outbreaks earlier. While no one SNP was identified among isolates with the same stress tolerant phenotype, increased sequencing of L. monocytogenes isolates in combination with stress tolerance profiling, will enhance our ability to identify genetic elements associated with more stress tolerant strains.    100  Chapter 3: Comparative analysis of Listeria monocytogenes plasmids and expression levels of plasmid-encoded genes during growth under salt and acid stress conditions  3.1 Introduction  Like many bacterial species, Listeria monocytogenes strains are known to harbour plasmids with frequencies reaching as high as 79% (Dykes et al., 1994; Fistrovici and Collins-Thompson, 1990; Kolstad et al., 1992; Lebrun et al., 1992; Perez-Diaz et al., 1982; Peterkin et al., 1992). To date, plasmids acquired by L. monocytogenes have been shown to contain genes that confer resistance to sanitizers,  (Elhanafi et al., 2010; Katharios-Lanwermeyer et al., 2012; Rakic-Martinez et al., 2011), heavy metals (Katharios-Lanwermeyer et al., 2012; Lebrun et al., 1992; Rakic-Martinez et al., 2011), and to a lesser extent, antibiotics (chloramphenicol, clindamycin, erythromycin, streptomycin and tetracycline) (Hadorn et al., 1993; Poyart-Salmeron et al., 1990). Listeria spp. plasmids also commonly contain several other uncharacterized efflux pumps including multiple drug resistance (MDR), small multidrug resistance (SMR) and multi antimicrobial extrusion (MATE) proteins (Boylan et al., 2006; Gibson et al., 2000; Kuroda and Tsuchiya, 2009; Masaoka et al., 2000). Additionally, a number of oxidative stress response genes (peroxidases, reductases) (Kuenne et al., 2010; Liang et al., 2016) have also been observed on L. monocytogenes plasmids but their exact roles have yet to be investigated. Interestingly, higher rates of plasmid harbourage have been reported among food and environmental isolates compared to clinical isolates (Lebrun et al., 1992; McLauchlin et al., 1997). Furthermore, plasmid harbourage rates have been shown to be higher amongst reoccurring strains  101  of L. monocytogenes (75%) as opposed to sporadic strains (35%) isolated from food or food-processing facilities (Harvey and Gilmour, 2001), suggesting that genes found on L. monocytogenes plasmids may be beneficial for the survival of this pathogen in such environments. This idea has in part already been confirmed by Elhanafi et al. (2010) who discovered a set of L. monocytogenes plasmid-encoded genes (bcrABC) responsible for increased resistance to benzalkonium chloride (MIC of 40 vs. 10 µg/ml for plasmid-cured strains), a quaternary ammonium-based disinfectant commonly employed by the food industry. However, such disinfectants are frequently applied at much higher levels (200 µg/ml) which efficiently eradicate L. monocytogenes. Nonetheless, frequent application of such disinfectants may create a selective pressure for L. monocytogenes isolates to uptake brcABC containing plasmids from other L. monocytogenes strains and thus lead to a higher prevalence among food-related isolates. A multitude of heavy metal (cadmium, copper, lead, zinc, mercury, arsenic) resistance genes have also been described on L. monocytogenes plasmids, though to date only plasmid-mediated cadmium resistance has been demonstrated (Lebrun et al., 1992; Maryse Lebrun et al., 1994; Xu et al., 2016). Such genes may be beneficial for survival in the environment as heavy metals have been found in both soil and water (Kasuya et al., 1992; Little and Martin, 1972; Ryan et al., 2000). Plasmid harbourage and subsequent replication poses an additional metabolic burden on cells and has been shown to lead to decreased growth rates relative to plasmid-cured strains [reviewed in Diaz Ricci and Hernández (2000)]. However, depending on the genes contained on a plasmid, plasmid-harbourage may also provide cells with a growth advantage when exposed to certain conditions. In Chapter 2, L. monocytogenes strains possessing a plasmid were found to have significantly faster growth rates in acidified media (BHIB pH 5) compared to plasmid-free strains, prompting the hypothesis that genes found on L. monocytogenes plasmids contribute to  102  acid tolerance. Furthermore, when plasmid-positive strains were compared based on their repA phylogeny (Kuenne et al., 2010), strains harbouring the larger group 2 plasmids had significantly faster growth rates in 6% NaCl than strains with the smaller group 1 plasmids. Given the importance of being able to control the growth of L. monocytogenes in foods, these findings warrant a deeper investigation into the roles that plasmids may play in the growth and survival of L. monocytogenes in foods.  The objectives of this study were to determine if genes found on L. monocytogenes plasmids contribute to the bacterium’s ability to survive food-related stresses (acid and salt) and to also assess the genetic commonalities and differences between repA sequence-based plasmid groups. 3.2 Materials and methods 3.2.1 Strains and culture conditions A total of seven L. monocytogenes strains were used in this study (Table 3-1. L. monocytogenes strains used in this study.  Five were plasmid-harbouring strains previously characterized in Chapter 2, and two were plasmid-cured strains graciously obtained from Michael Milillo (Milillo, 2015). Plasmid-cured strains were obtained via repeated passaging under increased temperature (45°C). The strains were stored at -80°C in brain heart infusion broth (BHIB, Difco, Fisher Scientific, Canada) with 20% glycerol and routinely cultured at 30°C on BHI agar (Difco, Fisher Scientific) plates.    103  Table 3-1. L. monocytogenes strains used in this study.  Strain Origin Serotype Genbank Accession number Plasmid typea Notable characteristics A58 Alberta, Canada 1/2b GCA_001711635.1 pLmG1-9 pLmG2-4  Acid tolerant  Desiccation tolerant   Contains two plasmids Lm10 British Columbia, Canada 1/2a GCA_001709685.1 pLmG2-9  Lm10_PC    N/A  Plasmid-cured Lm10 Lm20 British Columbia, Canada 1/2c GCA_001709785.1 pLmG1-11  Lm20_PC    N/A  Plasmid-cured Lm20 Lm106 British Columbia, Canada 1/2b GCA_001709455.1 pLmG1-12  Lm228 Switzerland 1/2b GCA_001709725.1 pLmG2-13  Salt tolerant  Acid tolerant  Contains the largest plasmid identified within our strain collection aG1 and G2 within the plasmid types denote the group that the plasmid belongs to as determined by a repA phylogeny. 3.2.2 Genetic characterization of plasmids A total of 93 concatenated plasmid sequences, previously described in Chapter 2, were used in this study. In short, all contigs of repA positive strains were aligned to the closed genome of Lm EDG-e using Mauve. Contigs not aligning to EGD-e were excluded as plasmid-associated if they contained chromosomal DNA elements or did not align to any previously published plasmids or other repA containing contigs. The resulting concatenated plasmid sequences were compared for genetic similarity by using Blastn and the sequences listed under BioProject  104  PRJNA329415 on the NCBI server (https://www.ncbi.nlm.nih.gov). The plasmids were also previously categorized as belonging to either group 1 (G1) or group 2 (G2) plasmids based on a repA phylogeny that was constructed according to Kuenne et al. (2010).  Plasmids sharing >99% nucleotide identity and differing by at most one mobile element gene, were categorized as belonging to a single plasmid type. One representative sequence of each plasmid type was extracted from the whole genome alignment of a strain, and annotated using ClassicRAST on The RAST v2.0 server (Aziz et al., 2008; Brettin et al., 2015; Overbeek et al., 2013) using the taxonomy ID 1639 for L. monocytogenes. The types of genes found on G1 and G2 plasmids were then compared.  FigTree v1.4.3 (http://tree.bio.ed.ac.uk/software/ figtree/) was used to visualize a Neighbour Joining phylogenetic guide tree of the various plasmid types that was produced using progressive mauve in Mauve v2.4.0 (Darling et al., 2010). For comparison, eight additional plasmid sequences (pLM80, pLM1-2bUG1, pLM5578, pCT100, pLM7UG1, pLm33, pLGUG1, pLI100) used in Kuenne et al. (2010) were also included in this analysis. Noteworthy, pLI100 is from an L. innocua strain and pLGUG1 is from an L. grayi strain while all others are from L. monocytogenes strains of human or food origin.  Plasmid types were also screened for antimicrobial resistance genes using ResFinder 2.1 available on the Center for Genomic Epidemiology, Technical University of Denmark server (https://cge.cbs.dtu.dk/services/ResFinder-2.1/) using 70% gene identity and 60% coverage as cutoffs.  105  3.2.3 RNA isolation and real-time qPCR analysis Three L. monocytogenes strains (A58, Lm106, Lm228) were used to evaluate the expression levels of plasmid encoded genes under salt and acid stress conditions. Lm106 and Lm228 were selected as representative G1 and G2 plasmid strains, respectively. Both contained a large number of genes (21-22 with functions) that were highly prevalent (>50%) among all plasmids belonging to these groups. A58 was also included because it uniquely contained both a G1 and a G2 plasmid and we were interested to know if having two plasmids may influence the bacterium’s ability to survive food-related stresses. One colony from each of the three L. monocytogenes strains was inoculated into 10 ml of BHIB and incubated at 30°C for 18 h. The cultures were then diluted to 107 CFU/ml in either fresh BHIB, BHIB+6% NaCl (w/w), or BHIB adjusted to pH 5 and again incubated at 30°C. At mid-exponential phase (108 CFU/ml, absorbance at 600 nm (A600nm) 0.08-0.1), cellular metabolism was halted by adding 10% phenol:cholorform (Fisher Scientific) in ethanol solution pre-chilled to -80°C in a 1:10 volume to the sample. The tubes were vortexed briefly and then centrifuged immediately for 10 min at 4,696×g and 0°C. The supernatants were removed and the resulting pellets were stored at -80°C. Four biological replicates were performed for each strain and treatment combination. Total RNA was isolated and purified using the PowerMicrobiome™ RNA Isolation kit (MO BIO Laboratories, CA, USA) per the manufacturer’s protocol. RNA integrity numbers (RINs) were determined using the 2100 Bioanalyzer (Agilent, CA, USA). Samples with a RIN between 9.7 and 10 were converted to cDNA using the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA) per the manufacturer’s protocol. Primers for select genes were designed using Primer3 Plus (Table B-1, Appendix B) and our draft whole genome sequences for the strains.  106  qPCR was conducted in a CFX96 TouchTM Real-Time PCR Detection System (BioRad) using SsoAdvanced TM Universal SYBR Green ® Supermix (BioRad). The thermocycling parameters used were as follows: initial denaturation for 30 s at 98°C, followed by 40 cycles of denaturation for 30 s at 95°C and elongation for 40 s at 57°C for 40 s. A melting curve was also performed at 65°C using 0.5°C increments. The relative expression levels of the plasmid encoded genes were calculated using the 2-ΔΔCT method (Livak and Schmittgen, 2001), with 16S rRNA as the reference gene (Tasara and Stephan, 2007). Genes with an average fold change > 2 or < -2 were considered statistically induced or suppressed, respectively. 3.2.4 Stress tolerance comparisons of wildtype and plasmid-cured strains Wildtype (Lm10, Lm20) and plasmid-cured strains (Lm10_PC, Lm20_PC) were compared on their ability to grow under salt and acid stress conditions using the protocols previously described in Chapter 2. In short, cultures were grown to stationary phase in BHIB at 30°C and then diluted in either BHIB+6% (w/w) NaCl or BHIB adjusted to pH 5 to achieve a final concentration of 107 CFU/ml. From these cultures, 200 µl was added in duplicate (technical replicates) to 96-well plates (Costar™ clear polystyrene, Fisher Scientific) that were incubated at 25°C in a microplate reader (Spectramax, V6.3; Molecular Devices, Sunnyvale, CA). The absorbance (A600nm) of each well was recorded every 30 min until the growth of all isolates reached stationary phase (~26 h) where absorbance readings begin to plateau. The resulting growth curves were fitted to the Baranyi and Roberts model (Baranyi and Roberts, 1994) using DMfit (v3.5) available on the ComBase browser (http://browser.combase.cc/DMFit.aspx). The entire experiment was repeated three times and model parameters of the three biological replicates were compared between the wildtype and PC strains using Student’s T-test with a 95% confidence level (p<0.05).  107  3.3 Results and discussion 3.3.1 Plasmid types and characteristics Among the 93 plasmid sequences, 26 unique plasmid types were detected with 13 belonging to repA G1 and 13 belonging to repA G2 (Tables 3-2 and 3-3). As previously mentioned in Chapter 2, G1 plasmids were significantly smaller (p<0.0005, 26-88 kbp) than G2 plasmids (55-107 kbp). Similarly, G1 plasmids contained 29-113 genes whereas G2 plasmids encoded 63-120 predicted genes. The GC contents of the two plasmid groups also differed slightly with G1 plasmids containing 34.4-36.9 GC, and G2 plasmids having slightly higher GC levels of 36.6-37.7%. These values are in line with what has been previously reported for L. monocytogenes plasmids (Liang et al., 2016; Schmitz-Esser et al., 2015) and are also consistent with the average GC content of L. monocytogenes genomes (37-38% GC) and other firmicutes such as Lactobacillus, Staphlylococcus, and Clostridium species (Bohlin et al., 2010; Kuenne et al., 2013).  It is interesting to note that while the G2 plasmid sequences consisted of one to three contigs, G1 plasmid sequences consisted of up to seven contigs with only three plasmids consisting of less than three contigs. Plasmid contigs were identical across all strains harbouring the same plasmid types. The number of contigs resulting from software assembled genomic sequences largely depends on the frequency of repeat elements such as prophages, transposases, and IS elements as these sections are often difficult to assemble (Edwards and Holt, 2013). Accordingly, it would appear that G1 plasmids contained a much larger number of mobile genetic elements (up to 18) compared to G2 plasmids (Table 3-2, Table 3-3), despite G2 plasmids being larger.  A complete sequence phylogeny of all plasmid types revealed three distinct groupings where repA G1 plasmids formed one large cluster with three outlying plasmid types, and G2 plasmids formed two discrete subgroups (Sub1 and Sub2, Fig 3-1). Within G2 Sub1 two separate  108  branches of similar plasmids types can also be seen (Fig 3-1, B1 and B2) with one of the G1 outlying plasmid types clustering with G2 Sub2 plasmids. The genetic similarities and differences between the clusters will be further discussed.     109  Table 3-2. Characteristics of G1 plasmids and their associated strains.  Plasmid type Sub-group Strains Origin Serotype CC Size (bp) Contigs GC ORFs MGEs pLMG1-1 MC 3 CH 1/2a, 1/2c (2) 9 25605 1 36.85 29 8 pLMG1-2 MC 1 CH 1/2c 9 38056 3 36.85 42 8 pLMG1-3 MC 1 AB 1/2a sing. 40730 3 36.85 48 6 pLMG1-4 MC 3 AB 1/2c 9 44350 3 36.85 54 11 pLMG1-5 MC 7 AB 1/2c 9 48409 (6), 48460 3 36.85 52 12 pLMG1-6 Outlier 4 BC 4b 6 54735 (3), 54736 5 34.77 64 9 pLMG1-7 MC 26 AB (18), BC (4), CH (4) 1/2a (17), 1/2b (6), 1/2c (2), 3b 3 (5), 5 (2), 7 (9), 9 (3), 11 (5), 89, 193 57076, 57082, 57083 (24) 3 36.04 64 13 pLMG1-8 MC 3 AB, BC (2) 1/2c 9 58105 3 36.65 69 13 pLMG1-9* MC 1 AB 1/2b 5 62258 2 35.58 75 17 pLMG1-10 MC 1 AB 1/2a/3a 199 70385 3 36.36 83 17 pLMG1-11 G2 Sub2 1 BC 1/2c 9 75351 7 36.85 88 13 pLMG1-12 MC 2 AB, BC 1/2b 5 78240, 78245 4 36.14 95 18 pLMG1-13 Outlier 5 BC 1/2a 11 87487 (2), 87488, 87574 (2) 1 34.38 113 15 * indicates one of two plasmids found in a single strain. MC = main cluster. Numbers in brackets denote the number of plasmids belonging to each characteristic when more than one is listed in a cell. Sing. refers to a singleton. Sequence size refers to the combined size of the assembled contigs contributing to each plasmid. All genes predicted to encode a mobile element protein, recombinase, transposase, integrase, invertase, or resolvase were considered to be mobile genetic elements (MGEs).        110  Table 3-3. Characteristics of G2 plasmids and their associated strains.  Plasmid type Sub-group Strains Origin Serotype CC Size (bp) Contigs GC ORFs MGEs pLMG2-1 Sub1-B2 1 AB 1/2a 8 55472 1 37.55 63 1 pLMG2-2 Sub1-B2 1 CH 1/2a 121 61053 1 36.85 64 4 pLMG2-3 Sub1-B2 6 BC 1/2a (4), 3a (2) 321 65637 2 36.85 74 6 pLMG2-4* Sub2 6 AB 1/2a 8 77109, 77221 (2), 77229 (3) 1 36.85 83 9 pLMG2-5 Sub1-B2 2 AB 1/2a 8 77249 2 36.85 83 10 pLMG2-6 Sub1-B1 2 AB 1/2c 9 81510 2 36.85 87 5 pLMG2-7 Sub2 3 AB, BC (2) 1/2a, 1/2b (2) 7, 88 (2) 81644 1 36.85 92 7 pLMG2-8 Sub2 1 AB 1/2b 5 87369 3 37.71 97 10 pLMG2-9 Sub2 1 BC 1/2a 155 89025 2 37.55 99 10 pLMG2-10 Sub2 4 AB, BC,  CH (2) 1/2a (2), 1/2b (2) 5 (2), 204 (2) 89996 2 36.85 99 11 pLMG2-11 Sub1-B1 6 AB (5), BC 1/2a 8 92204 (5), 99205 1 36.85 98 10 pLMG2-12 Sub1-B1 1 AB 1/2a 8 98358 2 36.85 108 12 pLMG2-13 Sub1-B1 1 CH 1/2b 59 107184 2 36.85 120 11 * indicates one of two plasmids found in a single strain. B1 and B2 refer to the two branches seen within G2 Subgroup 1 in Fig 3-1. Numbers in brackets denote the number of plasmids belonging to each characteristic when more than one is listed in a cell. All genes predicted to encode a mobile element protein, recombinase, transposase, integrase, invertase, or resolvase were considered to be mobile genetic elements (MGEs).    111    Figure 3-1. Neighbouring Joining phylogenic tree showing the genetic groupings of L. monocytogenes G1 and G2 plasmids. B1 and B2 refer to the two different plasmid lineages visible within G2 Subgroup 1.    112  3.3.2 Genetic elements shared by group 1 and group 2 plasmids A few features were identified among all or a large majority of the L. monocytogenes plasmids analyzed in this study. Table 3-4 lists all of the genes with predicted functions that are shared between G1 and G2 plasmids. Three genes were found on all plasmid types: repA, the plasmid replication protein; repB, the plasmid replication associated protein necessary for plasmid partitioning; and an excinuclease ABC subunit A involved in DNA repair. All three genes are located in the same region on the plasmids. In G1 plasmids the genes are separated from each other by one hypothetical protein (Fig 3-2 and Fig 3-3) while in the G2 plasmids there is no hypothetical protein between repA and repB. Other exceptions include plasmids pLMG1-11, pLMG2-11, and pLMG2-13 where the excinuclease is located elsewhere. It should also be noted that many plasmids contained several copies of the excinuclease. In a similar study by Kuenne et al. (2010), these same three genes were also identified on all 14 Listeria sp. plasmids investigated., L. monocytogenes chromosomes contain an excinuclease ABC subunit A gene (uvrA), but it does not share any genetic similarity to the plasmid encoded version.   The next most prevalent plasmid encoded genes were a cadmium-transporting ATPase and a cadmium efflux system accessory protein. These genes were found on all 13 G1 plasmid types and on 11 G2 plasmid types. However, two different versions of the genes existed among the plasmids. Eleven G1 and seven G2 plasmids contained the Tn5422 transposon which encodes both cadmium resistance genes and two transposases (Lebrun et al., 1994). This transposon is present on numerous Listeria sp. plasmid sequences available in the NCBI database. Kuenne et al. (2010) also reported that all but one of their plasmids (pLGUG1) contained two cadmium resistance genes. An additional two G1 and four G2 plasmid types contained different versions of the genes that were not followed by two transposases but by a hypothetical protein and a plasmid associated  113  integrase/recombinase. When aligned, the two different versions of the genes shared 98% gene coverage and 69% nucleotide identity. No cadmium resistance determinants were detected on pLMG2-1 or pLMG2-7. A number of plasmids from both G1 and G2 groups also contained a multi-metal transporter and a multicopper oxidase adjacent to the cadmium resistance genes (Fig 3-2, Fig 3-3). Although cadmium resistance genes can also be found on Listeria sp. chromosomes, they are most commonly found on plasmids (Lee et al., 2013). Only four of our 166 L. monocytogenes strains possessed chromosomally encoded cadmium resistance genes that when aligned with the Tn5422 plasmid-encoded cadmium resistance genes, shared 28% gene coverage and 75% nucleotide identity. Other genes belonging to both G1 and G2 plasmids encoded an NADH peroxidase, the L-proline glycine betaine ABC transport system permease protein ProW, and a lead, cadmium, zinc and mercury transporting ATPase. These three elements are located in the same region with NADH peroxidase and proW next to each other, followed by three hypothetical proteins and then the multi-metal transporter with mobile transposable elements flanking either end of the region. Plasmid pLMG1-4 differed in that it did not contain the multi-metal transporter. The genes found in this region have been previously reported in other L. monocytogenes plasmid studies (Kuenne et al., 2010; Liang et al., 2016).  Ten G1 plasmid types (2-8, 10-12) and seven G2 plasmid types (2-3, 5-6, 11-13) contained a ~8414 bp region that existed either as part of the repA containing contig, or as an individual separate contig (Fig 3-2, Fig 3-3). In all G2 plasmids the region was located on the repA contig. The region contains 10 complete ORFs and two additional ORFs on either end of the region that were incomplete in some plasmid types (Fig 3-2, Fig 3-3). When the complete versions existed,  114  they encoded a predicted DNA methylase and a hypothetical protein. Eight additional genes were annotated as hypothetical proteins and two encoded a DNA binding protein and a cell filamentation protein (Fic).  The arsenic resistance cassette previously identified in pLI100 (Kuenne et al. 2010), was found on two G1 and three G2 plasmid types in this study. Interestingly, all strains harbouring pLMG1-1, G1-2, G1-4, G1-5, and G1-8 plasmids, as well as three strains harbouring pLMG1-7 plasmids, contained the arsenic resistance cassette on a chromosomal contig. Four additional plasmid-free strains also contained this cassette. While L. monocytogenes chromosomes have been previously shown to harbour this arsenic resistance cassette (Lee et al., 2013), its presence has not yet been associated with strains harbouring specific plasmid types. The fact that all strains harbouring particular G1 plasmid types contained this cassette suggests that the cassette was likely at some point part of these plasmids and migrated into the chromosome. It is also possible that the four plasmid-free strains harbouring the arsenic resistance cassette might also have contained plasmids at one point.   Another notable observation from the plasmid comparison analysis was the presence of the PemIK toxin/antitoxin stable maintenance system in six G1 and six G2 plasmids (Fig 3-2, Fig 3-3). In the G1 plasmids these genes are located adjacent to the cadmium resistance cassette. This toxin-antitoxin system has been described for the Lactobacillus salivarius UCC118 plasmid pSF118–20 (Fang et al., 2008) and was also found to be present on many of the Listeria sp. plasmids analyzed in Kuenne et al. (2010). Other genes with a high prevalence between both repA plasmid groups included those encoding a clpL ATPase, DNA helicase HerA, DNA topoisomerase III, a number of transcription regulators, and a putative ATPase TraE which was found on all G2 plasmids and on pLMG1-13  115  which was not part of the major G1 plasmid cluster (Table 3-4). Consistent with previous literature, no antibiotic resistance genes were found on any of the plasmids. While L. monocytogenes plasmids have been found to contain genes that facilitate resistance to chloramphenicol, erythromycin, streptomycin, and tetracycline (Hadorn et al., 1993; Poyart-Salmeron et al., 1990), such incidences remain rare. More often, antibiotic resistance genes such as tetA, strA, and floR, are chromosomally located.     116  Table 3-4. Predicted proteins observed on both group 1 and group 2 plasmids. Rows highlighted in yellow represent proteins found on 10 or more plasmid types from one plasmid group. qPCR was performed on genes encoding the proteins highlighted in blue. Predicted protein G1 G2 Predicated protein G1 G2 ABC-type transport system ATPase 1 1 Multidrug resistance protein EbrA 2 5 Arsenate reductase 2 3 Multidrug resistance protein EbrB 2 5 Arsenical pump-driving ATPase 1 3 Na+/H+ antiporter 1 1 Arsenical resistance operon repressor 2 3 NADH peroxidase 9 8 Arsenical resistance operon trans-acting repressor ArsD 1 3 NADPH:quinone oxidoreductase 2 2 4 Arsenical-resistance protein ACR3 1 3 Negative transcriptional regulator-copper transport operon 2 5 ATP-binding protein p271 1 4 Ferroxidase 1 1 ATP-binding protease subunit clpL 8 6 PHP N-terminal domain protein 1 5 Bipolar DNA helicase HerA 5 5 Pli0005 protein 10 8 Cadmium efflux system accessory protein 13 11 Pli0006 protein 10 8 Cadmium-transporting ATPase 13 11 Pli0007 protein 11 9 ClpB protein 1 2 Pli0008 protein 11 8 conserved hypothetical protein - phage associated 8 6 Pli0010 protein 11 8 Copper chaperone 1 1 Pli0011 protein 11 8 Copper-translocating P-type ATPase  1 1 Pli0046 protein 8 8 CrcB protein 2 1 Pli0065 protein 2 4 DEAD/DEAH box helicase-like protein 1 5 Pli0073 protein 1 5 Death on curing protein, Doc toxin 6 6 Pli0074 protein 1 5 DNA methylase 2 12 Pli0075 protein 1 5 DNA topoisomerase III 4 8 Prevent host death protein, Phd antitoxin 6 6 DNA-directed RNA polymerase beta subunit  1 3 Prolipoprotein diacylglyceryl transferase  2 1 DNA-entry nuclease  1 1 Putative ATPase TraE 1 13 DNA-invertase 11 7 Replication initiation protein 13 13 DUF1541 domain-containing protein 4 4 Replication-associated protein 13 12        117  Predicted protein G1 G2 Predicated protein G1 G2 Enolase  2 2 Resolvase 5 5 Glyoxalase family protein 3 4 Resolvase X 2 4 Hypothetical SAV0801 homolog in superantigen-encoding pathogenicity islands SaPI 1 2 Resolvase/integrase Bin 4 5 ImpB/MucB/SamB family protein 13 13 Site-specific recombinase, resolvase family 9 7 Integrase/recombinase plasmid associated, putative 3 4 Tn552 transposase 1 4 Lead, cadmium, zinc and mercury transporting ATPase; Copper-translocating P-type ATPase  10 8 Tn554-related, transposase B 1 3 Lin1741 protein 6 6 Transcriptional regulator, Crp/Fnr family 1 1 L-proline glycine betaine ABC transport system permease protein ProW  9 8 Transcriptional regulator, TetR family 4 5 LtrC-like protein 2 5 Transposase 2 4 Mannose-6-phosphate isomerase  1 1 Transposase A from transposon Tn554 1 3 Mobile element protein 13 12 Transposase and inactivated derivatives-like protein 9 6 MoxR-like ATPase 1 4 Type III restriction enzyme, res subunit:DEAD box helicase 1 5 Multicopper oxidase 8 5 Type III restriction-modification system methylation subunit 1 5    YbbM seven transmembrane helix protein 1 1     118   Figure 3-2. Complete alignment of L. monocytogenes G1 plasmid sequences. The 13 G1 concatenated plasmid sequences from this study were aligned using Mauve. Four additional sequences from Kuenne et al. (2010) were included for comparison and are highlighted in red font. Regions filled in with the same colour share a high degree of genetic similarity. White boxes mark the locations of genes on the plasmids. It should be noted that Mauve was not able to identify all homologies due to algorithmic limitations.   119     Figure 3-3. Complete alignment of L. monocytogenes G2 plasmid sequences. The 13 G2 concatenated plasmid sequences from this study were aligned using Mauve. Four additional sequences from Kuenne et al. (2010) were included for comparison and are highlighted in red font. Regions filled in with the same colour share a high degree of genetic similarity. White boxes mark the locations of genes on the plasmids. It should be noted that Mauve was not able to identify all homologies due to algorithmic limitations.   120  3.3.3 Genetic elements associated with group 1 plasmids A total of 43 genes were uniquely observed on G1 plasmids, excluding those with hypothetical functions (Table 3-5). The most prevalent unique G1 genes were a CRISPR-associated protein and two hypothetical proteins from L. innocua which were found on five plasmid types all belonging to the main G1 cluster. Otherwise, most genes unique to G1 plasmids were only found on one or two plasmid types. A second toxin-antitoxin system containing a zeta toxin and epsilon antitoxin gene was found on two G1 plasmid types. This system has also been reported on streptococcal and enterococcal plasmids (Sletvold et al., 2008; Zielenkiewicz and Ceglowski, 2005). A third plasmid type additionally contained a RelE/StbE replicon stabilization toxin and a RelB/StbD replicon stabilization antitoxin system that has also been found on Escherichia coli, Bartonella ancashensis, and Klebsiella pneumoniae plasmids  among others (Hayes, 1998; Kaplan et al., 2015; Mullins et al., 2017).  Other notable genes found on G1 plasmids included those encoding alcohol dehydrogenase, glycerol dehydrogenase and glycerol kinase with collective roles in glycerol utilization; a copper-transporting ATPase; a CrcB protein which encodes a fluoride riboswitch which aids in removing fluoride from cells; a mercury resistance cassette; a MATE family multidrug resistance protein; a phosphate regulatory protein (PhoB); three protein subunits of a phosphoenolpyruvate-dependent dihydroxyacetone (Dha) kinase (DhaKLM), and an ATP-dependent Dha kinase. The alcohol and glycerol dehydrogenases and glycerol kinase existed adjacent to the genes encoding DhaKLM, two ATP-dependent Dha kinases, and a TetR transcription regulator. Mobile genetic elements flank either end of the region (Fig 3-2). TetR has been shown to activate the transcription of dhaKLM, and glycerol dehydrogenase and kinase are used to convert glycerol into Dha which is one of the simplest carbohydrates making it an  121  important precursor for the synthesis of organic compounds in bacteria (Christen et al., 2006; Erni et al., 2006). Strains harbouring plasmids with this region may have improved access to carbon which may be used for the production of energy or cellular constituents such as membrane phospholipids, which could potentially enhance the ability of a strain to tolerate certain stresses.    Overall, G1 plasmids harboured many of the same sets of genes but several individual plasmid types also had their own additional unique elements. Three G1 plasmid types differed greatly from the other G1 types with pLMG1-11 clustering with G2 plasmids. pLMG1-6 formed its own unique cluster with pCT100, an L. monocytogenes G1 plasmid from Kuenne et al. (2010). These two plasmids share a unique region that encodes an additional multi-metal transporter, a copper-transporting ATPase, a copper operon negative transcriptional regulator, and a MATE family multidrug resistance protein. pLMG1-13 did not group with any of the other plasmid types analyzed in this study with more than half of the plasmid containing unique genes which can be found in Table 3-5.    122  Table 3-5. Predicted proteins uniquely observed on group 1 plasmids. Proteins highlighted in yellow were found on five or more plasmids. Plasmid types highlighted in blue were not closely related to any other G1 plasmids analyzed in this study. Predicted protein G1 plasmid types Sum 1 2 3 4 5 6 7 8 9 10 11* 12 13 Acyltransferase family protein             1 1 Alcohol dehydrogenase          1   1  2 Arsenic efflux pump protein   1           1 ATPase involved in DNA repair             1 1 Cadmium resistance protein             1 1 Copper-transporting ATPase      1        1 CRISPR-associated protein MTH1087    1   1  1 1  1  5 Dihydroxyacetone kinase, ATP-dependent          2   2  2 DUF1706 domain-containing protein             1 1 Epsilon antitoxin         1   1  2 Glutathione-dependent formaldehyde dehydrogenase      1        1 Glycerol dehydrogenase          1   1  2 Glycerol kinase          1   1  2 Hypothetical lambda repressor-like, DNA-binding             1 1 Integral membrane protein      1        1 Lmo0466 protein    1   1  1 1  1  5 Lmo2276 protein    1   1  1 1  1  5 Membrane proteins related to metalloendopeptidases             1 1 Mercuric ion reductase           2    1 Mercuric resistance operon regulatory protein          2    1 Methyl-accepting chemotaxis protein           1   1 Multi antimicrobial extrusion (MATE) family transporter      1        1 Myosin heavy chain, nonmuscle type B             1 1 Na(+)/H(+) antiporter  1            1 Organomercurial lyase           1    1 Oxidoreductase (putative)         1   1  2 Permease of the drug/metabolite transporter superfamily      1        1 Phage protein             1 1 Phosphate regulon transcriptional regulatory protein PhoB       1        1  123  Predicted protein G1 plasmid types Sum 1 2 3 4 5 6 7 8 9 10 11* 12 13 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase ADP-binding subunit DhaL         1   1  2 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase dihydroxyacetone binding subunit DhaK         1   1  2 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase subunit DhaM         1   1  2 Predicted transcriptional regulator of pyridoxine metabolism      1        1 Prophage LambdaSa2, site-specific recombinase         1   1  2 Protein involved in cell division             1 1 Protoporphyrinogen IX oxidase, novel form, HemJ              1 1 pXO2-10             1 1 RelB/StbD replicon stabilization protein (antitoxin to RelE/StbE)             1 1 RelE/StbE replicon stabilization toxin             1 1 RepB             2 1 Replication-associated protein RepB         1   1  2 Rlx-like protein             1 1 Site-specific recombinase, DNA invertase   1   1       1 3 Site-specific recombinase, phage integrase family             1 1 Sortase A, LPXTG specific             1 1 Tn916, transcriptional regulator, putative         1   1  2 Transcriptional regulator, PadR family             1 1 Transcriptional regulator, XRE family             1 1 Transcriptional repressor, BlaI/MecI family      1        1 Transposase, IS204/IS1001/IS1096/IS1165      2   1   1  3 Type I restriction-modification system, restriction subunit R              1 1 Zeta toxin         2   2  2 The plasmid marked with a * was more closely related to G2-Sub2 plasmids. It should be noted that RAST was unable to annotate all genes successfully.    124  3.3.4 Genetic elements associated with group 2 plasmids On average G2 plasmids contained more genes than G1 plasmids (90 vs 67); however, there was less diversity among these genes, which is presumably associated with the lower number of mobile genetic elements on these plasmids. Unique to G2 plasmids was the presence of virulence associated genes, as well as a cell surface protein, a general secretion pathway E protein, and a membrane bound protease which were all found on eight or more G2 plasmid types (Table 3-6).  G2 plasmids formed two distinct subgroups (1 and 2) and within subgroup 1 two additional branches existed (Fig 3-1, B1 and B2). G2 clusters were found to be associated with plasmid size. Sub1-B2 contained the three smallest G2 plasmids and pLMG2-5; Sub1-B1 contained the three largest G2 plasmids as well as the much smaller plasmid pLMG2-6; and Sub2 contained plasmids of intermediate size (Table 3-2).  All G2 plasmids containing the 8414 bp region (n=8) that was found on both G1 and G2 plasmids, contained an adjacent region of approximately 18.5 kbp that was unique to G2 plasmids and contained genes encoding a lipoprotein, ATP TraE, DNA topoisomerase III, a membrane bound protease, and a Type IV secretory pathway (Fig 3-3). Noteworthy, no portion of the 18.5 kbp region was detected in any G2 Sub2 plasmids. With the exception of pLMG2-3, this region lay directly adjacent to another G2 specific region of ~12 kbp that encoded a general secretion pathway E protein and a cell surface protein (Fig 3-3). This region was prevalent among almost all G2 plasmid types with pLMG2-2 and pLMG2-3 being exceptions. However, the general secretion pathway E protein was retained in these plasmids despite the absence of the remaining parts of 12kbp region.   125  Within the G2 plasmid subgroups, many plasmid types shared a high degree of genetic similarity. Within Sub1-B2, pLMG2-1 and pLMG2-2 plasmids differed only by the presence of a ClpL ATPase and the cadmium resistance transposon (Tn5422) in pLMG2-2. Plasmids pLMG2-2 and pLMG2-3 differed only by the presence of a mobile genetic element containing the multidrug efflux pumps ebrA and ebrB, along with a TetR transcription regulator. ebrAB are responsible for resistance to ethidium, acriflavine, pyronine Y and safranin O in E. coli and Bacillus subtilis and belong to the SMR family of multidrug efflux pumps (Masaoka et al., 2000). Lastly, pLMG2-5 differed from the other Sub1-B2 plasmids by the presence of the three different regions containing the secretion pathway E and cell surface protein, NADH peroxidase and proW, and a multicopper oxidase and multi-metal transporter. Within Sub1-B1, pLMG2-13 differed from G2-11 and G2-12 by the presence of a region containing a DNA topoisomerase III, a mannose-6-phosphate isomerase, a YbbM seven transmembrane helix protein, ferroxidase, prolipoprotein diacyclglyceryl transferase, and a copper chaperone and copper ATPase. This region is downstream of the multicopper oxidase and multi-metal transporter genes in pLMG2-13 and shows no similarities with any of the other G2 plasmids (Fig 3-3). This same region is also found on the G1 plasmid pLMG1-8. pLMG2-11 and G2-12 differ only by the presence of clpB encoding a stress induced chaperone protein, and the region containing NADH peroxidase and proW. Similarly, the only notable difference between pLMG2-6 and pLMG2-10 was the presence of the arsenic resistance cassette in pLMG2-10. Additionally, the pLMG2-6 plasmid from our study was found to be identical to pLM5578 from Kuenne et al. (2010).   126  It is interesting that both the largest and smallest plasmids were closely related while the intermediate sized plasmids formed their own very distinct cluster far from both G1 and all other G2 plasmids. It is possible that the large plasmids derived from the small plasmids through the accumulation of transposable elements. Plasmids from Sub2 are unique in that they all contain a ~43 kbp region that encodes a large number of genes, in particular many associated with virulence: VirB4 and VirD4 secretion system proteins, an invasion protein, a thermonuclease, DNA methylase, a DEAD box helicase, and a type III restriction enzyme and modification system. Moreover, plasmids G2-4, G2-8, G2-9, G2-10, and pLM80 all contained an additional region encoding the multidrug resistance proteins EbrA and EbrB.    127  Table 3-6. Predicted proteins uniquely observed on group 2 plasmids. Proteins highlighted in yellow were found on five or more plasmids. qPCR was performed on genes encoding the proteins highlighted in blue. Red plasmid types belong to G2 Sub1-B1 and red plasmid types belong to G2 Sub1-B2. Remaining plasmid types belong to G2 Sub2.  Predicted protein G2 plasmid types Sum 1 2 3 4 5 6 7 8 9 10 11 12 13 ABC transporter           1   1 Cell surface protein 1   1 1 1 1 1 1 1 1 1 1 11 Chromosome (plasmid) partitioning protein ParA 1             1 Conjugation protein, TraG/TraD family, (pXO2-16) 1 1 1           3 Conserved hypothetical protein    1   1 1 1 1    5 General secretion pathway protein E 1 1 1 1 1 1 1 1 1 1 1 1 1 13 Hypothetical protein, (pXO1-65) 1   1 1 1 1 1 1 1 1 1 1 11 Hypothetical protein, (pXO2-28) 1 1 1 1 1 1 1 1 1 1 1 1 1 13 Invasion associated protein p60    1   1  1 1    4 Lipoprotein, NLP/P60 family 1 1 1  1 1     1 1 1 8 Membrane protein, putative, (pXO2-14)    1   1 1 1 1    5 Membrane-bound protease, CAAX family 1 1 1  1 1     1 1 1 8 Phage protein lin1266    1    1 1 1    4 Pli0009 protein    1   1 1 1 1    5 Pli0068 protein  1 1 1 1 1  1 1 1 1 1 1 11 Secreted antigen GbpB/SagA/PcsB, putative peptidoglycan hydrolase       1       1 Thermonuclease     1   1 1 1 1    5 Tn5252, Orf 21 protein, internal deletion    1    1  1    3 TolA protein 1   1    1 1 1    5 TraG/TraD family protein    1   1 1 1 1    5 Type IV secretory pathway, VirD4 components     1 1     1 1 1 5 Type V secretory pathway, adhesin AidA 1 1 1   1        4 It should be noted that RAST was unable to annotate all genes successfully. 128  3.3.5 Expression of L. monocytogenes plasmid encoded genes during growth under salt and acid stress conditions To investigate the hypothesis that plasmids contribute to enhanced salt and acid tolerance in L. monocytogenes, the expression levels of a number of plasmid encoded genes were investigated in three L. monocytogenes strains during growth in BHIB with 6% NaCl or pH 5. The investigated genes were selected based on being prevalent on both G1 and G2 plasmids, or being uniquely present in a high number of plasmid types from a single group. Furthermore, genes were chosen based on having functions with probable roles in bacterial stress response.  In strain Lm106 the pLMG1-12 genes encoding NADH peroxidase, cadmium-transporting ATPase, ProW, ClpL, and multicopper oxidase (MCO) were all upregulated >2-fold during growth in 6% NaCl (Figure 3-4). Similar results were seen for growth at pH 5 with the exception of proW which had a <2-fold level of induction. Near identical expression changes were observed under both stress conditions for the same set of genes found on either pLMG1-9 or pLMG2-4 within strain A58. A58 contained two cadmium ATPases (one on each plasmid), and while both were upregulated during growth in 6% NaCl and at pH 5, the pLMG2-4 version was activated to a lower extent compared to the pLMG1-9 version.  NADH peroxidases have been proven necessary for the decomposition of hydrogen peroxide which accumulates in cells during aerobic growth (Gibson et al., 2000; Sakamoto and Komagata, 1996), and may also have roles in bacterial virulence (La Carbona et al., 2007). Broadly speaking, NADH peroxidases aid in mediating oxidative stress which occurs in bacteria exposed to a number of different stresses due to the disruption of cellular homeostasis which leads to enhanced levels of reactive oxygen species (Mittler, 2002). In this study, plasmid-encoded NADH  129  peroxidases appear to be beneficial for growth under salt and acid stress conditions with higher induction levels observed under acid stress as compared to salt stress (Fig 3-4).  proW encodes an L-proline glycine betaine ABC transport system permease protein. Both L-proline and glycine betaine are compatible solutes that have been shown to accumulate in multiple bacteria and plants during osmotic stress exposure, usually through membrane transportation as opposed to de novo synthesis (Gerhardt et al., 2000; Khedr et al., 2003; Ko et al., 1994; Ko and Smith, 1999; Sleator and Hill, 2002; Yoshiba et al., 1997). These solutes are believed to help stabilize proteins against denaturation as well as counteract the outflow of water from cells under hypertonic growth conditions (Kempf and Bremer, 1998). It is, therefore, not surprising that a plasmid-encoded osmolyte transporter was upregulated in L. monocytogenes subjected to salt stress (Fig 3-4). The uptake of compatible solutes has not previously been associated with acid stress and our findings also support this. On the contrary, Clp ATPases have been long associated with bacterial acid stress responses and virulence (Frees et al., 2007; Gaillot et al., 2000; Varcamonti and Arsenijevic, 2006; Wall et al., 2007). These enzymes are widely conserved bacterial chaperone proteins which have critical roles in refolding and degrading damaged cell proteins (Wall et al., 2007).  A plasmid-encoded multicopper oxidase was also upregulated in L. monocytogenes during growth in 6% NaCl and at pH 5. Aside from the proven role of these enzymes in copper homeostasis, they also show enhanced oxidase activity for a wide range of substrates and participate in transmembrane iron transport (Askwith et al., 1994; Huston and Jennings, 2002). Since iron homeostasis and responses to oxidative stress are coordinated (Cornelis et al., 2011), multicopper oxidase may have a role in mediating the oxidative stress imposed by salt and acid conditions. Like many of the genes found on L. monocytogenes plasmids, multicopper oxidases  130  have not been found on L. monocytogenes chromosomes but are made available to this pathogen through plasmid harbourage. The roles of cadmium ATPases  in cadmium detoxification and resistance have been thoroughly studied in several bacteria (Katharios-Lanwermeyer et al., 2012; Maryse Lebrun et al., 1994; Nucifora et al., 1989), but their putative functions in bacterial salt and acid tolerance remain unknown. Casey et al. (2014) found that a plasmid-encoded cadmium ATPase was upregulated 2.45-fold in L. monocytogenes cells exposed to 4 ppm of benzethonium chloride, a quaternary ammonium based sanitizer. It is therefore possible that like multicopper oxidase, cadmium ATPases have the ability to act on additional substrates. However, unlike multicopper oxidase, cadmium ATPases are also found within transposons embedded in L. monocytogenes chromosomes (Lebrun et al., 1994).  Three G2 plasmid specific genes encoding a cell surface protein, a general secretion pathway protein E, and a multidrug efflux pump (ebrA) were downregulated in strain A58 during growth in both 6% NaCl, and pH 5 media. However, in Lm228 different versions of a cell surface protein and general secretion pathway protein E showed increased expression under salt stress and decreased expression under acid stress. Cell surface proteins participate in a broad range of activities including environmental signaling, surface and cell adhesion, and pathogenesis (Bierne and Cossart, 2007; Navarre and Schneewind, 1999). The secretion pathway protein E on the other hand, shared 100% protein identity with a type IV secretion protein which is typically involved in triggering a host response that promotes virulence in L. monocytogenes (Woodward et al., 2010). The putative role of this protein in salt tolerance remains to be discovered.  Multidrug efflux pumps, such as that encoded by ebrA investigated in this study, have been shown to have roles in bacterial virulence, as well as bile and sanitizer resistance (Elhanafi  131  et al., 2010; Lin et al., 2003; Romanova et al., 2006). However, despite bile containing high levels of salt, we did not observe the induction of this gene during the growth of L. monocytogenes in salt or acid media (Fig 3-4).   Similar to Lm106 and A58, Lm228 also showed >2-fold increased expression of NADH peroxidase and cadmium ATPase under both stresses and the glycine betaine transporter was again only significantly (>2-fold) upregulated under salt stress. The pLMG2-13 plasmid in this strain additionally contained a membrane-bound protease and a lipoprotein which were common among G2 plasmids (Table 3-6). Both genes were upregulated under salt stress but showed decreased or little induction under acid stress. The lipoprotein shares 99% protein identity to multiple L. monocytogenes cell wall hydrolases which aid in the growth and development of cell walls and can be major determinants of growth rate and cell wall architecture (Lee and Huang, 2013). The pLMG2-13 protease shares 99% protein identity with a intramembrane metalloprotease which serves to maintain homeostasis of cell surface components by cleaving damaged or misfolded proteins that can occur following a stress (Kroos and Akiyama, 2013).  The excinuclease ABC subunit A (uvrA) gene did not show any significant induction in any of the three strains under the stresses evaluated, but consistently had low levels of induction (1.6-1.9 fold) under salt stress and was downregulated under acid stress. This protein was selected for use in our expression analysis because the same version of the gene was present in all L. monocytogenes plasmids evaluated in this study. This enzyme is involved in DNA repair, specifically DNA damage caused by UV radiation, oxidative stress, and bile salts (Kim et al., 2006; Payne et al., 2013; Prieto et al., 2006) with the latter likely explaining why it was consistently induced at low levels in all strains exposed to 6% NaCl.   132  Collectively, the results demonstrate that a number of genes both commonly and uniquely found on L. monocytogenes plasmid types are activated in the pathogen during growth under salt and acid stress conditions. Furthermore, while the focus of this research is on plasmid-mediated survival in food-related environments, both salt and acid stress conditions are also frequently encountered along the human gastrointestinal tract suggesting that plasmid harbourage may also assist gastrointestinal passage, and thereby enhance the virulence potential of a strain. It is also interesting to note that in order for genes to appear suppressed under a specific set of conditions, they must show decreased expression relative to a control condition. This demonstrates that the gene is expressed under control conditions thus suggesting that plasmid-encoded genes may also have fundamental roles in supporting normal cell functions which may further encourage plasmid maintenance.   133   Figure 3-4. Differential expression of L. monocytogenes plasmid-encoded genes in strains subjected to salt (BHIB+6% NaCl, 30°C) and acid stress (BHIB pH 5, 30°C) in comparison to the control (BHIB, 30°C). RNA was extracted from mid-exponential phase cells. Bar heights represent the average log2 fold changes and error bars denote standard deviations (n=4). BHIB = brain heart infusion broth; MCO = multicopper oxidase; Cd = cadmium, Secret path = secretion pathway E.   134  3.3.6 Stress tolerance comparisons of wildtype and plasmid-cured strains To determine if the induction of plasmid encoded genes is responsible for the enhanced growth rates of plasmid-harbouring strains in pH 5 media, and for the increased salt tolerance of G2 plasmid strains relative to G1 plasmid strains, the growth rates of two plasmid-cured strains (Lm10_PC, Lm20_PC) were compared to those of their wildtype strains (Lm10, Lm20). When grown in BHIB at 25°C, both Lm10 and Lm20 exhibited maximum growth rates and cell densities that were similar (p>0.05) to those of the plasmid-cured strains (Fig 3-5). However, as seen in Fig 3-5, the plasmid-cured strains exhibited longer lag phase durations (~1 h longer, p≤0.001) compared to the wildtype strains. This is the opposite of what was hypothesized, as plasmid harbourage is often thought, and has been shown by some studies, to pose an additional burden on cells during replication (Diaz Ricci and Hernández, 2000; Feng et al., 2013; Goverde et al., 1994). However, in support of our findings Logue et al. (2006) also found that a plasmid-cured strain of Yersinia enterocolitica had a longer lag-phase duration than its wildtype strain when grown in BHIB at 25°C, but that the exponential growth rates were comparable. This again may suggest that plasmid encoded genes have active roles in normal cell growth, in particular the processes involved in transitioning from lag phase to exponential growth. When the wildtype and plasmid-cured strains were grown in 6% NaCl and pH 5 media the results were similar to those observed for growth in BHIB. In 6% NaCl, the plasmid-cured strains had lag phase durations that were significantly longer (3-4 h, p≤0.012) than those obtained for the wildtype strains (Fig 3-5). In pH 5 media the lag phase duration of Lm20_PC was significantly (p=0.050) longer by 2 h compared to Lm20, while no significant difference was found between Lm10_PC and Lm10 (Fig 3-5). Again, the maximum growth rates and maximum cell densities did not differ (p>0.05) between plasmid-cured and wildtype strains in 6% NaCl or at pH 5. Based on  135  these findings, it appears that plasmid harbourage may be more beneficial for growth in 6% NaCl than at pH 5. However, since the plasmid-cured strains exhibited longer growth phases in BHIB as well, it is difficult to conclude that this difference was not just more visible under salt stress conditions. In Chapter 2, no significant difference (p>0.05) was found between the salt tolerance of plasmid-positive vs. plasmid-negative strains; however, it’s possible that any growth advantages were masked in the strains by other confounding factors such as the absence of genes with roles in salt tolerance or disadvantageous variations of such genes.  While Lm10 contained pLMG2-9 and Lm20 contained pLMG1-11, both belonged to G2 Sub2 shown in Fig 3-1, and likewise contained similar genetic elements. Both plasmids contained proW, NADH peroxidase, cadmium-transporting ATPase and ebrA genes which were used in our expression experiment. Lm10 additionally contained the same versions of the general secretion pathway E and cell surface proteins found on pLMG2-4 in strain A58. Although the results do not completely support the findings from Chapter 2 where plasmid harbouring strains had faster growth rates in pH 5 media compared to other naturally occurring plasmid-free strains, these strains only represent two of the 26 different plasmid types identified. Furthermore, these plasmids were not found in any other L. monocytogenes strains in our collection and unlike most other plasmids, they did not contain any toxin-antitoxin systems which together could indicate that these plasmids are newly introduced into L. monocytogenes or not stable.     136   Figure 3-5. Growth of wildtype L. monocytogenes strains and their plasmid-cured (PC) counterparts in BHIB, BHIB +6% NaCl, and BHIB pH 5 at 25°C. A) Lm10 and Lm10_PC, B) Lm20 and Lm20_PC. Data points denote the averages of three replicates and error bars represent standard deviations.   137  3.4 Conclusions This study examined both the genetic similarities and differences between two L. monocytogenes repA sequence-based plasmid groups (G1 and G2) and the putative roles of specific plasmid-encoded genes in the tolerance of L. monocytogenes to salt and acid stress. Our results showed that G1 and G2 plasmids contained many of the same genes but also many that were unique. Overall, G1 plasmids had a larger diversity of genes compared to G2 plasmids, despite G2 plasmids being larger and containing more genes. While G1 plasmids formed one main phylogenic cluster, G2 plasmids formed two distinct subgroups with subgroup 2 plasmids containing multidrug resistance and numerous virulence associated genes. Across all L. monocytogenes plasmids was an abundance of genes with putative roles in bacterial stress response. qPCR revealed that a number of these genes are activated in L. monocytogenes during growth under salt and acid stress conditions. Specifically, a NADH peroxidase, cadmium ATPase, multicopper oxidase, and ClpL chaperone protein were all upregulated >2-fold during growth in 6% NaCl and pH 5 media. Furthermore, an osmolyte transporter ProW was upregulated >2-fold during growth in 6% NaCl media, and a lipoprotein specific to G2 plasmids was upregulated >2-fold during growth in pH 5 media. A comparison of the growth rates of two plasmid-cured strains to their wildtype strains during growth under both stresses did not reveal any differences in growth rates between the strains, but demonstrated longer lag phases for the plasmid-cured strains in all conditions tested including the control. Collectively, this work suggests that L. monocytogenes plasmids may have roles in facilitating survival under salt and acid stress conditions. This would pose an additional concern for the food industry where food-related L. monocytogenes isolates have already been shown to have higher rates of plasmid harbourage. The presence of a plasmid may help L. monocytogenes  138  survive longer in food-processing environments and also decrease the lag phase duration of strains in preserved ready-to-eat foods. Furthermore, plasmid harbourage may also be associated with improved gastrointestinal survival as salt and acid stress conditions are frequently encountered by L. monocytogenes along this route. The detection of several virulence associated genes also warrants a further investigation into how plasmid harbourage may be influencing L. monocytogenes virulence.   Strains from Canada and Switzerland harboured identical plasmids containing genes involved in tolerance to salt and acid stress, suggesting that efforts should be made to limit the spread of these plasmids that may be occurring in the global food market. It is reasonable to expect that plasmids that support the growth and survival of L. monocytogenes will continue to spread in the population. As whole genome sequencing is becoming increasingly more affordable and popular, an emphasis should be placed on screening L. monocytogenes sequences for plasmids and making global comparisons to determine their spread. This will allow us to not only be able to track more easily the evolution of L. monocytogenes plasmids, but also monitor the overall prevalence of plasmid-harbourage among food-related isolates. Lastly, the majority of predicted plasmid encoded genes remain hypothetical proteins, emphasizing how much we still have to learn about plasmids and their possible impacts on bacteria.   139  Chapter 4: Strand specific RNA-sequencing and membrane lipid profiling reveals growth-phase dependent cold-stress response mechanisms in Listeria monocytogenes  4.1 Introduction  The human pathogen Listeria monocytogenes continues to pose a challenge in the food industry, where it is known to contaminate ready-to-eat foods and to grow during refrigerated storage. Ingestion of L. monocytogenes by susceptible individuals can cause potentially fatal food-borne infections with mortality rates as high as 40% reported (Ramaswamy et al., 2007). The ubiquitous nature of this pathogen makes it difficult to eliminate from our food systems; however, post-processing levels of L. monocytogenes contamination are often low (Cabedo et al., 2008; Fenlon et al., 1996; Kozak et al., 1996) and unlikely to cause disease (Buchanan et al., 1997; Chen et al., 2003). Furthermore, the risks associated with L. monocytogenes contamination can be addressed by heating foods to an appropriate temperature (Doyle et al., 2001). Therefore, the ability of L. monocytogenes to grow to unsafe levels in ready-to-eat foods during the shelf-life of products, poses the greatest concern to consumer health.  When subjected to low temperatures, all living cells will experience similar challenges that stem from a reduction in molecular dynamics, which leads to decreased rates of diffusion and structural changes to molecular structures (Thieringer et al., 1998). Cold-stress adaptation mechanisms among microbes include membrane compositional changes, compatible solute uptake, and the synthesis of nucleic acid stabilization proteins, DNA-unwinding enzymes, and general stress response and cold shock proteins (Phadtare, 2004; Weber and Marahiel, 2003).  140  Currently, most of our knowledge regarding bacterial cold-stress response (CSR) mechanisms comes from the model organisms Escherichia coli and Bacillus subtilis. However, unlike L. monocytogenes, neither of these bacteria can multiply at temperatures close to 0°C. In the last decade, L. monocytogenes outbreaks associated with ready-to-eat foods in both North America and Europe have increased interest in the specific mechanisms employed by this psychrotrophic pathogen to adapt and grow at low temperatures.  The CSR of L. monocytogenes has been studied using microarrays (Chan et al., 2007b; Cordero et al., 2016; Durack et al., 2013), quantitative real-time PCR (qPCR) (Arguedas-Villa et al., 2010; Chan et al., 2007a; Markkula et al., 2012; Mattila et al., 2012; Pieta et al., 2014), mutagenesis (Angelidis et al., 2002; Becker et al., 2000; Borezee et al., 2000; Chan et al., 2008; Chassaing and Auvray, 2007; Liu et al., 2006; Zhu et al., 2005a) and other genetic and proteomic techniques (Bayles and Wilkinson, 2000; Cacace et al., 2010; Liu et al., 2002). Collectively, these methods have identified a large pool of genes with putative or known roles in cold tolerance. However, although a small portion of qPCR-based studies have looked at the differential expression (DE) of select genes during the early stages of the L. monocytogenes CSR (e.g., lag phase) (Arguedas-Villa et al., 2010; Chan et al., 2007a), global transcriptome studies have only been conducted on cold-adapted exponential- and stationary-phase cells to date, leaving much to be discovered regarding the initial CSR of L. monocytogenes. Furthermore, since CSRs become more pronounced in response to more dramatic changes in temperature, many studies have focused on abrupt temperature down shifts from 37°C to 15°C or lower (Arguedas-Villa et al., 2010; Chan et al., 2007b; Markkula et al., 2012; Mattila et al., 2012; Pieta et al., 2014). Such conditions may not represent the food industry, where more commonly L. monocytogenes is transferred from an ambient-temperature environment to a food product or plant environment that is maintained at 4 -  141  10°C. Though exposure to ambient temperature (20-25°C) can also be considered a low temperature stress (Knudsen et al., 2004; Markkula et al., 2012), the maximum growth rate (μmax) of L. monocytogenes is approximately 10× faster at 20°C than at 4°C whereas there is much less of a difference between 15 °C and 37°C (~4.5×) (Baranyi and Tamplin, 2004). Moreover, at 37°C, the transcriptional landscape of L. monocytogenes undergoes a drastic change in preparation for intracellular survival (Toledo-Arana et al., 2009).  Recent advances in molecular and sequencing technologies now allow us to detect several forms of non-coding RNA (ncRNA). Studies of ncRNA have increased our understanding of gene regulation and opened a new area in bacterial stress response research. ncRNAs exist in several different forms and thus participate in a wide range of functions. Most ncRNAs can be divided into three main categories: 1) Cis-regulatory RNAs, 2) trans-encoded small RNAs (sRNAs), and 3) antisense RNAs (asRNAs) (Mellin and Cossart, 2012). Cis-regulatory RNAs are located at the 5’-ends of mRNA and fold into alternative structures in response to physicochemical cues. These transcripts are often referred to as riboswitches or thermosensors. One of the best-known examples in L. monocytogenes is the riboswitch that blocks the translation of the major virulence regulator prfA at low temperatures (<30°C) (Loh et al., 2009). Trans-encoded sRNAs, on the other hand, are not located adjacent to their target and share only limited complementarity allowing them to regulate multiple mRNAs. Lastly, asRNAs are transcribed from the DNA strand opposite of a gene and thus have perfect complementarity. asRNAs can be short (<100 nt) or long (>1000 nt) and in some cases correspond to the 5’- or 3’- extension of an mRNA transcribed from an adjacent gene (Thomason and Storz, 2010). The prevalence of genes with reported asRNA ranges from 20-75% depending on the organism (Mraheil et al., 2011; Sharma et al., 2010). To date, more than 100 asRNAs have been described  142  in L. monocytogenes (Behrens et al., 2014; Mraheil et al., 2011; Toledo-Arana et al., 2009; Wehner et al., 2014). Compared with the numbers reported for other bacteria this number seems rather low. However, it will presumably rise with further study.  The objective of this study was to gain a comprehensive view of the L. monocytogenes CSR, with an emphasis on elucidating mRNA and asRNA expression patterns across multiple growth phases following cold stress. To gain a better understanding of the mechanisms employed by L. monocytogenes to grow during refrigerated storage, we aimed to characterize the CSR from cells during the initial lag phase following cold stress, throughout exponential growth and long-term stability at low temperatures, using conditions similar to those present in a food-contamination scenario. To accomplish these goals, strand-specific RNA sequencing and membrane lipid profiling were conducted on L. monocytogenes cell cultures at five distinct growth phases, following a downshift in temperature from 20°C to 4°C. As the phenotypic and genetic properties of L. monocytogenes isolates differ, we chose to elucidate the CSR mechanisms of a strain with relevance to the food industry and consumer health: a fast cold-growing, serotype 1/2a, food plant isolate containing full length versions of important virulence genes: inlA, inlB, inlC, prfA, plcA, hly, mpl, actA, plcB. Here, we report novel findings regarding cold-induced membrane compositional changes in L. monocytogenes, the growth-phase dependent cold-stress regulon, and the active roles of antisense transcripts in regulating the CSR of this pathogen.   143  4.2 Materials and methods 4.2.1 Culture conditions and time point selection A previously evaluated L. monocytogenes food-related isolate (BioSample SAMN05256775) from Chapter 2 (Lm1, serotype 1/2a, sequence type 7) was selected for use in this study. The strain displayed enhanced cold tolerance (µmax at 4°C) relative to a large collection of isolates. Bacterial cultures were grown for 24 h in brain heart infusion broth (BHIB; Difco, Fisher Scientific, Canada) at 20°C, re-suspended in pre-tempered BHIB at a cell density of 107 CFU/ml and incubated at 4 or 20°C. RNA and lipids were extracted at five time points from cells grown at 4°C (treatment, T) and 20°C (control, C). Three biological replicates were conducted for each treatment and control sample. Each time point corresponded to a specific growth phase (G) (Fig 4-1). The time points were as follows: G1 – early lag phase, G2 – transition to exponential growth phase, G3 – mid-exponential growth phase, G4 – transition to stationary phase, and G5 – late-stationary phase. G1 corresponded to 20% of the lag phase duration at each temperature. The timing was determined by modelling previously obtained growth curve data using the Baranyi and Roberts model (Baranyi and Roberts, 1994) in DMfit (v3.5) (http://browser.combase.cc/DMFit.aspx) (Baranyi and Tamplin, 2004). G2 was marked by a doubling in cell numbers, confirmed using plate counts. G3 corresponded to mid-exponential growth (108 CFU/ml), and G4 corresponded to the transition to stationary phase; both time points were identified spectrophotometrically (A600nm) and confirmed with plate counts. Lastly, G5 was selected to correspond to G4 plus 8× the lag phase duration (h) at each temperature.   144   Figure 4-1. Cell growth phase sampling strategy at 4°C and 20°C. 4.2.2 Fatty acid analysis At each of the five growth phases for cells grown at both 20°C and 4°C, cultures were pelleted (10 - 45 mg net weight), rinsed twice with one ml phosphate buffered saline (PBS, Fisher Scientific), and stored at -80°C. One sample per time point was collected. The frozen pellets were later sent to MIDI labs (Microbial ID, Inc., Newark DE, USA) where the cell lipids were extracted, followed by methylation of the fatty acids (FAs), and then loaded onto a gas chromatograph for analysis. Fatty acid methyl ester (FAME) profiles were then generated and analyzed using Sherlock® pattern recognition software.  4.2.3 RNA isolation and sequencing At each sample point, growth in cultures was stopped by adding 10% phenol:chloroform (Fisher Scientific) in ethanol solution pre-chilled to -80°C in a 1:10 volume to the sample. The tubes were vortexed briefly and then centrifuged immediately for 10 min at 4,696×g and 0°C. The supernatants were removed and the resulting pellets were stored at -80°C for less than 2 weeks.  145  Total RNA was isolated and purified using the PowerMicrobiome™ RNA Isolation kit (MO BIO Laboratories, CA, USA) per the manufacturer’s protocol. RNA integrity numbers (RINs) were determined using the 2100 Bioanalyzer (Agilent, CA, USA). Samples with a RIN between 9.7 and 10 and an RNA concentration >100 ng/μl were sent to Genome Québec (Montréal, QC, Canada) for rRNA-depleted Illumina TruSeq RNA library prep and TruSeq stranded total RNA 100 bp paired-end sequencing on the Illumina HiSeq 2000 platform. 4.2.4 RNA-seq data analysis Sequencing quality was assessed using FastQC (Andrews 2010), and Illumina adapter sequences and low-quality base pairs were removed using default parameters in Trimmomatic v 0.36 (Bolger et al., 2014). After the removal of low-quality reads, 22 - 39 million reads remained for each dataset. Reads were mapped to the complete sequenced genome of L. monocytogenes EGD (NCBI RefSeq NC_022568.1) using Bowtie 2 v 2.2.6 (Langmead and Salzberg, 2012) and allowing zero mismatches. L. monocytogenes EGD was selected as the reference genome as it allowed for the highest percent of successfully mapped reads compared to the more commonly employed reference strains EGDe and 10403S. The mapping efficiency ranged from 97.8 - 99.6% for individual reads and 92.1 - 95.6% for successfully paired reads. BAM alignment files were used as input for read counting using featureCounts v 1.5.0-p1 (Liao et al., 2014). The default counting mode ‘union’ was used, and separate count files were generated for sense and antisense transcripts. DE analyses were performed using DESeq2 (Love et al., 2014) in R v 3.1.1 (R Core Team, 2016), and the DE was reported as log2 fold changes. p-values were adjusted by the DESeq2 default Benjamini-Hochberg adjustment method and genes with a >2-fold (>1 log2) change in expression and an adjusted p-value < 0.05 were considered as DE. A principal component analysis  146  (PCA) was conducted using DESeq2 to determine the overall reproducibility of our RNA-seq biological replicates. To evaluate the overall transcription levels for individual genes at both 4°C and 20°C, raw read counts were normalized based on library size. No p-values were calculated for genes with low mean normalized counts, zero counts, or extreme outlier counts.  4.2.5 Clustering of gene expression profiles Clustering of genes with similar expression patterns across the five growth phases, was performed using the Mfuzz package (Kumar and E Futschik, 2007)(Kumar and E Futschik 2007) in R. Default filtering and standardization methods (based on standard deviation) were applied to the log2 data from all five growth phases. Soft clustering was then performed using the fuzzy c-means algorithm and the following parameters: c=20, m=1.25. Genes were assigned to clusters of given expression patterns across the growth phases based on having a membership value >0.5.  4.2.6 Functional categorization of differentially expressed genes To investigate the roles of DE genes at each growth phase, we determined the overrepresentation of molecular and biological pathways (Fisher exact test, p<0.05) using SmartTables based on the BioCyc database (https://biocyc.org/) (Caspi et al., 2016). SmartTables were also used to map L. monocytogenes EGD genes to the equivalent genes in L. monocytogenes 10403S for which the BioCyc database contains transcription regulation and gene ontology information. Additional enrichment analyses were then performed (Fisher’s exact test, p<0.05) for gene ontology (GO) terms, and genes regulated by certain transcription regulators (e.g., σB, PrfA, RpoD, VirR, MogR, CodY, HrcA and CtsR).  147  4.2.7 Quantitative PCR validation of RNA-seq data RNA-seq results were validated using qPCR amplification of three genes: 1) leuA which exhibited >4-fold higher expression at 4°C at all five growth phases, 2) cspB which exhibited >4-fold lower expression at 4°C at all growth phases, and 3) recJ which was chosen as a reference gene because it showed very little variation in expression across all growth phases at either temperature (Supplenmentary 4-2). Up to 1 µg of RNA from each of our 30 samples was reverse transcribed using the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA) per the manufacturer’s protocol. Primers were designed using Primer3 Plus (Table 4-1) and our draft whole genome sequence for Lm1 (GenBank Accession number GCA_001709805.1). qPCR was conducted in a CFX96 TouchTM Real-Time PCR Detection System (BioRad) using SsoAdvanced TM Universal SYBR Green ® Supermix (BioRad). The relative expression levels of leuA and cspB were calculated using the 2-ΔΔCT method (Livak and Schmittgen, 2001), with recJ as the reference gene. Table 4-1. Primers used for quantitative PCR validation of RNA-seq data. Gene abbr. Primer sequence Opt. Anneal. Temp. Source cspB Fw-CAAACAGGTACAGTTAAATGGTTTA 55°C (Arguedas-Villa et al., 2010)   Rv-ACGATTTCAAATTCAACGCTTTGA   leuA Fw-TTGTCGGGTATGCCTGTTCC 55°C This study  Rv-GGGTTTTTCAGCACGCCATC   recJ Fw-CTCGACCGGCAATTGTGTTG 55°C This study  Rv- GTCCACACTTCGACCAGACC      148  4.2.8 Accession numbers FastQ files of HiSeq runs were deposited into the NCBI Sequence Read Archive under BioProject PRJNA384077. 4.3 Results and discussion 4.3.1 mRNA transcriptome of the L. monocytogenes cold-stress response A total of 11 - 17 million paired-end mRNA reads per sample were successfully assigned to L. monocytogenes EGD open reading frames (ORFs). The number of reads mapped to each EGD ORF ranged from 0 to 630,000. No counts were observed for 27 ORFs, of which, 26 were confirmed to be absent in our strain, leaving one gene (LMON_0476) with no detectable expression at 20°C or 4°C. Overall, >99% of EGD ORFs were expressed by Lm1 at 20 and 4°C. This percentage is in line with the findings of Toledo-Arana et al. (2009), who reported that Listeria spp. express more than 98% of their ORFs at 30°C and 37°C.  Three of the 30 sequenced samples were excluded from further analysis as they were deemed to be outliers as visualized in the PCA plot in Fig 4-2. The excluded samples included one biological replicate each from the T4, C4, and C5 treatments. Two of the three samples were from G4 which represents the transition from exponential growth to stationary phase. Given the short window of time available for extracting RNA from this very specific physiological phase, it is likely that these samples were processed either too early or too late relative to the other two biological replicates. The excluded C5 sample belonged to the same biological replicate as the excluded C4 sample. The PCA plot (Fig 4-2) also shows that the level of transcriptome variance is much greater (67%) among growth phases than between the two temperatures (9%), highlighting the importance of growth phase selection in determining the outcomes of an experiment.   149  Overall, 1.3× more genes were suppressed than induced in L. monocytogenes under cold stress (Fig 4-3). This finding contrasts with the results reported in a microarray study conducted by Chan et al. (2007b), where 30-40% fewer genes were downregulated than upregulated in L. monocytogenes at 4°C compared to 37°C. However, in agreement with their work, we observed that cold-adapted stationary-phase cells exhibited the greatest number of DE genes (G5, Fig 4-3). From here on, the DE of genes at 4°C vs. 20°C will be discussed with respect to the growth phases (G1-G5) and C1-C5 and T1-T5 will be reserved for when referring to overall levels of transcription or changes in the membrane lipid profiles that occur at 20°C and 4°C, respectively. Fig 4-4 showed the pairs of growth phases that shared the highest and lowest numbers of the same up or downregulated genes at 4°C relative to 20°C. G4 had the lowest number of upregulated genes in common with the other growth phases, whereas G1 and G2 (G1:G2) and G1:G5 shared many of the same upregulated genes. With respect to downregulated genes, G4 and G5 had the smallest in common with G2 and G3 whereas G1:G2 and G1:G5 shared a high number of downregulated genes.  To confirm our DE findings, we used RT-qPCR to quantify the expression levels of a gene that was upregulated for >4 fold in response to cold at all growth phases (cspB), and a gene that was downregulated for >4 fold in response to cold at all growth phases (leuA), according to the RNA-seq data. We observed a strong positive correlation (R2=0.96, y=1.20x-0.41) between the mRNA levels detected by the two methods. Nevertheless, the fold changes calculated using RNA-seq data were consistently larger than those obtained using RT-qPCR (Fig C-1, Appendix C).  150   Figure 4-2. Principal component analysis plot of RNA sequencing biological replicates. C1 through C5 refer to the L. monocytogenes control cultures grown at 20°C, and T1 through T5 represent the treated cultures grown at 4°C (see Fig 4-1 for information about the sampling points). Samples marked with * were excluded from our analyses due to an abundance of outlier data points.      151   Figure 4-3. Numbers of L. monocytogenes sense and antisense RNAs differentially expressed at five growth phases in response to cold stress (4°C relative to 20°C). Differential expression (DE) analyses were performed using DESeq2 and DE was reported as log2 fold changes. Genes with a >2-fold change (> 1 log2) in expression and an adjusted p-value < 0.05 were considered DE. See Fig 4-1 for information about growth phases.  152   Figure 4-4. Heatmaps showing the number of L. monocytogenes sense and antisense RNA co-upregulated or co-downregulated (>2-fold) between pairs of growth phases at 4°C. Numbers outside of the pyramids represent the number of genes uniquely upregulated or downregulated at each growth phase. See Fig 4-1 for information about growth phases. 4.3.2 Core set of genes induced by cold stress  The DE analyses revealed a core set of 22 genes induced at 4°C relative to 20°C at all five growth phases (Table 4-2). These genes fall into the category of cold acclimation proteins (CAPs), which are defined as proteins encoded by genes that show continuous increased expression throughout prolonged cold-stress exposure (Hebraud and Guzzo, 2000; Thieringer et al., 1998) . Among these CAPs were nine genes involved in the production of branched-chain amino acids (BCAAs) (ilvDBHC-leuABCD-ilvA). These BCAA synthesis genes were statistically overrepresented among the upregulated genes at all growth phases (Table 4-3). The BCAAs leucine, isoleucine, and valine, are critical for BCFA production, with isoleucine being the  153  precursor for anteiso BCFAs. Listeria and similar bacteria, such as Bacillus species, alter their membrane lipid composition in response to temperature downshifts, to include higher percentages of branched-chain fatty acids (BCFAs) with anteiso configurations (Annous et al., 1997; Kaneda, 1991; Suutari and Laakso, 1992; Zhu et al., 2005a). Anteiso BCFAs have lower melting points than their iso-branched and straight-chain counterparts. They are thus more effective at increasing membrane fluidity, which is needed to maintain membrane-transport functions at low temperatures.  Other genes with increased expression in response to cold stress at all growth phases, included those encoding the L-carnitine transporter OpuC (opuCB-opuCC-opuCD), internalins A and D, a ribosomal protein (RpmF), a histidine biosynthesis protein (HisE), a GNAT family acetyltransferase (LMON_0625), and Veg protein (LMON_0187). The roles of carnitine and ribosomal proteins in bacterial adaptation to certain stresses have been previously reported (Angelidis et al., 2002; Gardan et al., 2003; Schmalisch et al., 2002; Wemekamp-Kamphuis et al., 2004b). In B. subtilis Veg protein has been shown to activate extracellular matrix genes and contribute to biofilm formation (Lei et al., 2013). In L. monocytogenes, increased expression of Veg has been observed following exposure to acid and cold stress (Balachandran et al., 2012; Chan et al., 2007b), but its function remains unknown.  In addition to the upregulation of inlA and inlD at all growth phases at 4°C, inlH and two genes encoding internalin-like proteins (LMON_2407, LMON_0611) were upregulated at four growth phases (Table 4-2). The most recognized roles of internalins A and B are in L. monocytogenes virulence. Nevertheless, the Listeria spp. family of internalin proteins all share a leucine-rich-repeat domain that allows them to bind structurally unrelated ligands, thereby implicating them in a wide range of functions (Kobe and Kajava, 2001). In Chapter 2, isolates  154  containing a full-length version of inlA exhibited enhanced cold tolerance relative to those with a truncated version. Further research is needed to confirm the potential role(s) of internalins in the L. monocytogenes CSR. Many more genes (n=104) were upregulated at four growth phases at 4°C relative to 20°C, with the majority being induced at G1, G2, G3, and G5 (G1-G3:G5, n=70). These genes shared similar DE patterns across all growth phases (Fig 4-5, Cluster 5 and 12), where levels of DE dramatically decrease at G4. Present in this group were genes involved in arginine biosynthesis (dapE, argGHJ), general stress response (LMON_0515, LMON_2234, LMON_2696, LMON_2770), FA synthesis (LMON_0351), and phosphotransferase system (PTS) uptake of cellobiose (LMON_2800-2803) and mannose (LMON_0785-0788) (Table 4-2). Additionally, the genes encoding succinate-semialdehyde dehydrogenase (LMON_0920), 11 genes from the pentose phosphate pathway (LMON_0350-0360), and two sets of genes (LMON_0354-0356, LMON_2718-2720) encoding dihydroxyacetone kinase (DhaKLM) were upregulated in these growth phases (Table 4-2).  An additional 100 genes displayed induced transcription at three growth phases, with the largest number co-upregulated at G1 - G2:G5 (n=29), followed by G1 - G3 (n=24). The similarities between these sets of growth phases are visible in Fig 4-4. At G1 - G2:G5, L. monocytogenes growth is inhibited, and such conditions would be expected to induce genes encoding proteins associated with 1) secondary membrane-transport systems that are less energy demanding, 2) the utilization of alternative energy sources, and 3) cell detoxification. It appears that the demand for these proteins is further increased in cold stressed cells as we observed the upregulation of genes involved in utilizing ethanolamine as an alternative energy source (LMON_1167-1180), oligopeptides transport (oppC), transcription regulation (5 genes), general stress response (clpC),  155  and ribosomal subunit assembly (LMON_2523). At G1 - G3, several of the shared upregulated genes encoded reductases (LMON_1496, LMON_2403, LMON_0798, LMON_0643, LMON_2843) with known or putative roles in maintaining redox balance within the cell and managing oxidative stress. Table 4-2. Core set of genes upregulated (>2-fold) across multiple (≥4) growth phases in L. monocytogenes cells at 4°C vs. 20°C. EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_0187 Veg protein  + lmo0189 16/14 2.96 2.45 2.28 2.38 1.54 LMON_0209 LSU ribosomal protein L25p ctc + lmo0211 10 2.16 3.43 2.72 0.87 1.58 LMON_0260 Internalin H  inlH + lmo0263 12 2.21 3.67 2.11 -1.54 1.57 LMON_0261 Internalin D  inlD +  9 1.86 2.79 2.18 1.23 2.70 LMON_0350 Transaldolase   lmo0343 5 1.88 2.03 1.56 0.27 4.16 LMON_0351 3-oxoacyl-[acyl-carrier protein] reductase   lmo0344 5 2.35 1.71 1.30 0.62 4.34 LMON_0352 Ribose 5-phosphate isomerase B   lmo0345 5 2.75 2.30 1.25 0.26 4.61 LMON_0353 Triosephosphate isomerase   lmo0346 5 2.38 1.95 1.75 0.45 4.38 LMON_0354 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase, ADP-binding subunit DhaL   lmo0347 5 2.21 2.08 1.70 0.03 4.21 LMON_0355 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase, dihydroxyacetone binding subunit DhaK   lmo0348 5 2.34 1.85 1.60 0.26 4.50 LMON_0356 Hypothetical protein   lmo0349 5 1.94 2.17 1.35 -0.16 5.37 LMON_0357 Hypothetical protein   lmo0350 5 3.01 1.62 1.45 -0.25 5.55 LMON_0358 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase, subunit DhaM   lmo0351 5 2.49 1.79 1.57 -0.37 5.50 LMON_0360 Acetyltransferase, GNAT family  + lmo0353 12 1.85 1.50 1.23 0.67 1.72 LMON_0441 Internalin A  inlA + lmo0433 5 2.01 1.82 1.92 1.62 2.30 LMON_0486 LSU ribosomal protein L32p rpmF + lmo0486 19 1.43 1.46 1.76 1.75 1.22 LMON_0515 Universal stress protein  + lmo0515 12 1.34 1.76 1.12 -2.06 1.98 LMON_0561 Phosphoribosyl-ATP pyrophosphatase hisE  lmo0561 2 1.70 1.64 1.01 1.53 2.98 LMON_0611 Internalin-like protein   lmo0610 12 2.50 2.85 2.06 -0.73 2.19 LMON_0625 Acetyltransferase, GNAT family   lmo0624 5 2.07 2.19 1.67 1.33 4.46 LMON_0626 Hypothetical protein   lmo0625 5 2.18 2.19 1.59 1.23 4.43 LMON_0785 PTS system, mannose-specific IID component   lmo0781 12 1.81 2.76 1.32 -0.54 1.97  156  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_0786 PTS system, mannose/fructose-specific IIC component    lmo0782 12 1.60 2.72 1.43 -0.76 1.64 LMON_0787 PTS system, mannose-specific IIB component   lmo0783 12 1.86 2.80 1.43 -1.11 1.86 LMON_0788 PTS system, mannose-specific IIB component / IIA component   lmo0784 12 2.05 3.10 1.49 -1.23 1.70 LMON_0920 Succinate-semialdehyde dehydrogenase [NAD]   + lmo0913 10 4.64 5.74 2.96 0.40 2.35 LMON_0944 Hypothetical protein    10 2.77 3.84 4.22 1.39 2.00 LMON_1488 Osmotically activated L-carnitine/choline ABC transporter, permease protein OpuCD opuCD  lmo1425 10 2.34 5.37 3.52 1.79 1.29 LMON_1489 Osmotically activated L-carnitine/choline ABC transporter, substrate-binding protein OpuCC opuCC  lmo1426 10 2.45 5.23 3.78 1.82 1.24 LMON_1490 Osmotically activated L-carnitine/choline ABC transporter, permease protein OpuCB opuCB  lmo1427 10 2.51 5.28 3.78 1.58 1.10 LMON_1491 Osmotically activated L-carnitine/choline ABC transporter, ATP-binding protein OpuCA opuCA  lmo1428 10 2.63 5.54 3.94 1.30 -0.01 LMON_1932 Predicted membrane protein hemolysin III  + lmo1864 14 2.21 1.58 1.78 1.84 1.06 LMON_2054 Dihydroxy-acid dehydratase ilvD + lmo1983 4* 3.52 3.89 3.34 4.29 3.62 LMON_2055 Acetolactate synthase large subunit ilvB + lmo1984 3/15 3.67 4.17 3.41 4.41 4.37 LMON_2056 Acetolactate synthase small subunit ilvH + lmo1985 15/1 3.67 4.02 3.39 4.41 4.03 LMON_2057 Ketol-acid reductoisomerase ilvC + lmo1986 4 2.98 4.08 3.00 3.96 3.28 LMON_2058 2-isopropylmalate synthase leuA + lmo1987 11 3.41 3.68 2.74 3.56 3.87 LMON_2059 3-isopropylmalate dehydrogenase leuB + lmo1988 3 3.18 3.60 2.86 3.84 4.72 LMON_2060 3-isopropylmalate dehydratase large subunit leuC + lmo1989 3 3.29 3.48 2.72 3.72 5.25 LMON_2061 3-isopropylmalate dehydratase small subunit leuD + lmo1990 3 3.16 3.64 2.71 3.96 5.70 LMON_2062 Threonine dehydratase ilvA + lmo1991 3 3.16 3.65 2.50 3.86 5.58 LMON_2234 Hypothetical protein   lmo2158 12 3.80 5.25 3.18 0.77 3.03 LMON_2407 Internalin-like protein   + lmo2396 3 1.13 1.52 0.92 1.30 2.67 LMON_2696 Universal stress protein  + lmo2673 12 3.09 3.64 1.90 -0.77 4.24 LMON_2718 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase, dihydroxyacetone binding subunit DhaK  + lmo2695 12 2.73 3.82 2.25 -0.43 2.36 LMON_2719 Phosphoenolpyruvate-dihydroxyacetone  + lmo2696 12 2.76 3.76 2.13 0.31 2.61  157  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 phosphotransferase , ADP-binding subunit DhaL LMON_2720 Phosphoenolpyruvate-dihydroxyacetone phosphotransferase , subunit DhaM;   + lmo2697 12 2.46 3.79 2.13 0.58 2.82 LMON_2735 Secreted protein    lmo2713 1 2.10 1.16 1.12 1.69 2.03 LMON_2770 general stress protein 26   lmo2748 12 2.51 3.72 1.71 -1.81 1.98 LMON_2800 PTS system, cellobiose-specific IIA component   lmo2780 12 1.79 1.57 1.45 -0.55 2.57 LMON_2801 Beta-glucosidase   lmo2781 12 2.06 1.38 1.45 -1.18 2.27 LMON_2802 PTS system, cellobiose-specific IIB component   lmo2782 12 2.11 1.49 1.81 -1.48 2.44 LMON_2803 PTS system, cellobiose-specific IIC component   lmo2783 12 2.32 1.24 1.80 -1.99 2.05 * indicates <0.50 cluster membership; G1-G5 refer to differential expression in L. monocytogenes cells grown at 4°C relative to 20°C and across five specific growth phases (see Fig 4-1); Blue shading indicates genes with ignificantly increased (> 1log2, p<0.05) gene expression and yellow shading indicates genes with significantly decreased (< -1log2, p<0.05) expression at 4°C relative to 20°C; Bolded values indicate significant (p<0.05) differential expression changes.                           158  Table 4-3. Pathways and gene ontology processes significantly (p<0.05*) enriched among genes upregulated (>2-fold) at 4°C vs. 20°C.  Biological process or pathway Gene examples p-value (# of contributing genes) G1 G2 G3 G4 G5 Branched-chain amino acid synthesis ilvABCDH, leuABCD, 1.18E-6 (9) 8.04E-8 (9) 9.14E-8 (9) 7.46E-11 (9) 5.75E-4 (9) Arginine biosynthesis argBDGHJ, carbB 3.42E-2 (4) 4.77E-4 (6) 4.25E-5 (7)  4.37E-3(8) Glycerol degradation LMON_0354-355, 0358, 2718-2720 2.87E-2 (6) 1.41E-5 (8) 9.82E-7 (9)  4.06E-3 (9) Histidine biosynthesis hisABDEGIZ 5.58E-5 (9) 7.20E-3 (4)  5.99E-9 (7) 1.16E-2 (6) Methionine biosynthesis metX, LMON_0595, 0847  4.99E-3 (3)  3.57E-2 (3)  Oxidation-reduction process (GO:0055114) qoxABCD, gabD  4.37E-2 (4) 1.54E-6 (8)  2.46E-4 (9) Response to stress (GO:0006950) recF, hrcA, grpE, dnaK, ltrC, uspA, csbD, ctsR, clpP, clpE 1.96E-5 (13)     Biological regulation (GO:0065007) yycF, ctsR, hrcA, frvA, opuCD 2.89E-3 (10)     Pyrimidine ribonucleotide biosynthesis pyrABC, carB    1.24E-3 (4)  Pentose phosphate pathway LMON_2682-2685, 2697, 0349, 0350, 0352     1.23E-3 (13) Carboxylates degradation fruA, bvrB, LMON_0095-0097, 2706-2708, 0785-0788     7.60E-6 (44) Tryptophan biosynthesis trpBCDFG)     2.96E-2 (5) Purine ribonucleotide biosynthesis purMQDFN     2.96E-2 (5) Active transporter activity (GO:0022804) opuCD, mpoABCD, mptACD, PTS systems     1.56E-8 (42) Structural constituent of ribosome (GO:0003735) rplEFNOPVX,, rpsEHNP, rpmC     2.90E-2 (20) * Statistical overrepresentations of gene sets were determined using Fisher’s exact test and the BioCyc database.    159   Figure 4-5. Differential gene expression patterns observed in L. monocytogenes cells grown across five growth phases at 4°C (see Fig 4-1). Clusters were formed using fuzzy c-means soft clustering and standardized log2 fold change values for 4°C expression levels vs. 20°C expression levels at each growth phase. Numbers in parentheses denote the number of genes in each cluster core. Represented genes had cluster membership values >0.5.    160  4.3.3 Core set of genes suppressed under cold stress  A total of 42 genes were downregulated at 4°C in all growth phases, including genes involved in cold shock (cspB), virulence (actA, plcB, capA, mpl), nucleotide degradation (LMON_0129, LMON_1234, guaB), vitamin B9 and B12 synthesis (LMON_1163, LMON_2771-2772), glycerophospholipid export (LMON_0106-0107), and sugar uptake (LMON_0770-0772, LMON_1451-1452). Many of the genes suppressed at 4°C contributed to either cobalamin (B12) biosynthesis and subsequent utilization, or the production and export of cell wall components (Table 4-4).  Cobalamin derivatives function as essential cofactors for important enzymes that catalyze a variety of transmethylation and rearrangement reactions (Matthews, 2001). In bacteria, cobalamin is necessary to degrade ethanolamine, propanediol, glycerol, and pyruvate and to produce carbon and energy (Garsin, 2010; Jeter, 1990; Matthews, 2001). L. monocytogenes and other bacteria induce the transcription of both cobalamin biosynthesis and ethanolamine utilization genes under a variety of stress conditions (Bowman et al., 2008; Jeter et al., 1984; Joseph et al., 2008, 2006; Joseph and Goebel, 2007). However, in the present study, we observed the downregulation of cobalamin biosynthesis genes and others encoding cobalamin-dependent enzymes and pathways (eutB, LMON_1144-1154, LMON_1803, LMON_0942), whereas the ethanolamine utilization operon remained upregulated at several growth phases.  Genes involved in the biosynthesis and export of peptidoglycan, teichoic acids, cell wall proteins, and membrane lipids were also suppressed at 4°C (Table 4-4). Both peptidoglycan and teichoic acid biosynthesis stem from the precursor molecule UDP-N-acetyl-α-D-glucosamine (UDP-GlcNAc) and require the attachment of D-alanine residues, and for peptidoglycan also glutamate. Correspondingly, genes involved in UDP-GlcNAc biosynthesis and transport, alanine  161  attachment (dltABCD), and glutamate transport (LMON_0849-0850) and attachment (racE) were downregulated. Other suppressed genes included several fatty-acid-coA ligases, lipid export proteins, and enoyl-acyl-carrier proteins and reductases, which collectively assist in the production and export of phospholipids, cell wall-associated lipoproteins, and lipoteichoic acids. These findings suggest a reduced rate of peptidoglycan/cell envelope turnover at 4°C relative to 20°C. This reduced turnover rate may reflect the reduced cellular growth rate at this temperature, or conservation of carbohydrates for alternative uses.  Other genes suppressed at 4°C had roles in localization and pathogenicity, heme biosynthesis, tRNA processing and charging, nucleotide degradation and salvage, and carbohydrate transport and catabolism (Table 4-4). Not surprising, many of these genes also have alternative roles in cell wall biogenesis. Notably, over 50 genes from PTS operons were downregulated at one or more growth phases, far exceeding the number of PTS genes upregulated in response to cold stress. PTSs utilize phosphate to facilitate the uptake of simple sugars and thus consume more energy than other membrane kinases with the same sugar specificity (Meadow et al., 1990). This suggests that L. monocytogenes may benefit from employing alternative sugar uptake systems when exposed to cold stress to conserve energy for more critical CSR mechanisms.       162  Table 4-4. Pathways and gene ontology processes significantly (p<0.05*) enriched among genes downregulated (>2-fold) at 4°C vs. 20°C. Biological process or pathway Gene examples p-value (# of contributing genes) G1 G2 G3 G4 G5 Cobalamin biosynthesis cbiKA, cbiP 5.82E-4 (5)  5.14E-6 (6) 4.01E-3 (6) 4.33E-4 (5) Biosynthesis of cell wall macrostructures (GO:0070589) dltABCD, pgdA, murABCE, pgi, tagD 5.76E-4 (6) 3.94E-4 (5) 1.67E-3 (5)  1.02E-4 (20) Pathogenesis inlK, inlJ, actA, arcB, flaA, ctaP 1.13E-2 (9) 4.83E-3 (7) 2.62E-2 (7)   Pyrimidine nucleotides/UMP biosynthesis pyrCBFGHR, comEB, udk, carB 2.60E-3 (8) 1.64E-6 (9) 3.03E-6 (11)   Localization (GO:0051179) dtpT, phoU, pstB, ctaP 1.03E-2 (13) 2.52E-3 (10)  4.59E-2 (12)  Transport (GO:006810) ctpA, pstB, phoU 1.03E-2 (13) 2.52E-3 (10)  4.59E-2 (12)  Carboxylate degradation PdhABC  1.11E-4 (21)  3.92E-8 (43)  UDP biosynthesis murABCE, racE, ddl 3.32E-3 (6)    2.74E-2 (6) tRNA charging metS, pheS, serS, tyrS, aspS, alaS, argS 1.80E-2 (8)     Myo-inositol degradation LMON_2235, 2238-2239    1.54E-2 (4)  Nucleotide degradation pdp, dra, drm, pnp, guaB    1.29E-2 (10)  Biosynthesis folD, cbiP, gatAC, comEB, hemABCD, cycEK, tagD, cdsA     5.50E-5 (128) Cofactors and electron carrier biosynthesis hemABCD, menEF, nadE, cbiP, ubiE     2.20E-3 (40) * Statistical overrepresentation of gene sets were determined using Fisher’s exact test and the BioCyc database.  4.3.4 L. monocytogenes cold-stress response at individual growth phases 4.3.4.1 Early lag phase (G1)  Previous studies have described the lag phase following cold stress as the acclimatization phase. In this phase, bacteria suppress bulk protein synthesis, and dramatically increase production of transient cold induced stress response proteins (CIPs) (Thieringer et al., 1998). Once the cells are cold adapted, CIP synthesis ends and bulk protein synthesis and cell growth resume. Depending on the degree of cold stress, the types and expression levels of CIPs will differ. Some of the  163  functions associated with CIPs include DNA unwinding, nucleic acid stabilization, and compatible solute uptake (Hunger et al., 2006; Jiang et al., 1997). A total of 96 genes were uniquely induced at G1, accounting for 25% of the genes upregulated at this growth phase. Most of these genes had DE patterns that belonged to clusters 14, 16, or 20 (Fig 4-5), with a large number encoding transcription regulators (n=19), including HrcA, CtsR, DegU, and YycF. Both HrcA and CtsR are negative regulators of several heat-shock and general stress-response genes, a number of which were also upregulated at G1 (grpE, dnaK, clpB, gmpA, mscAB, clpC, clpE, and clpP). Many of these upregulated genes, including groEL and groES, encode chaperone proteases capable of degrading the misfolded or damaged proteins that would probably occur following a rapid temperature downshift. By contrast, the transcription regulator DegU contributes to motility, growth at high temperatures, and efficient biofilm formation (Gueriri et al., 2008). Co-transcribed with degU is yviA, which encodes a DegV family protein with putative functions in FA transport or metabolism (Gueriri et al., 2008; Schulze‐Gahmen et al., 2003). As we did not observe upregulation of genes known to be activated by DegU, the role of this operon in early cold adaptation may be primarily associated with the functions of YviA. In support of this hypothesis, YycF, a transcription regulator of FA biosynthesis and cell wall metabolism genes (Howell et al., 2003), was also upregulated at G1.  Additional genes upregulated at G1 encode proteins involved in DNA repair (recF, gyrAB, recX, LMON_0230, LMON_1225), ethanolamine utilization (eutABCM, LMON_1166-1180), carnitine uptake (opuCABCD), phosphate transport and metabolism (pstABCS, phoU, phoB), RNA stability (LMON_0869, LMON_1037), nitrogen utilization (LMON_1582-1583), and histidine and arginine biosynthesis. Two genes with >20-fold increased expression at G1 encode succinate-semialdehyde dehydrogenase [NAD] (LMON_0920) and pyruvate phosphate dikinase  164  (LMON_1936) (Table 4-5). These proteins function in energy production, and glycerol metabolism, respectively.  Twenty-nine genes were solely induced at G1 and G2, highlighting their importance in the earlier CSR stages of L. monocytogenes. Included among these genes were those encoding the low temperature requirement protein LtrC, cold-shock protein CspL, regulatory protein RecX, and several proteins associated with phosphate transport (PstCABB-PhoU-PstS).  Forty-seven genes were upregulated at both G1 and G5, many of which followed the DE pattern shown in cluster 16 (Fig 4-5). As discussed, these growth phases both represent stages in cold stress survival, where the population size is static and their induced genes are probably specific to surviving adverse conditions. Among these genes were those encoding heat-shock proteins (clpP, clpE, groEL, LMON_2710-2711), DNA gyrase (gyrA), an SOS-response protein (lexA), an RNA helicase (secA), and seven transcription regulators. Interestingly, although these genes were upregulated in response to cold stress in this study, several of them were also reported by van der Veen et al. (2007) as induced in L. monocytogenes following 3 min of heat stress (48°C). As both stresses are caused by a rapid change in temperature which causes molecular structures to take on different conformations; it makes sense they would share similarly upregulated genes with functions in DNA and RNA repair (gyrA, lexA, secA), and protein degradation (clpP, clpE) or stabilization (groEL).     165  Table 4-5. Top 10 most highly induced genes in L. monocytogenes cells at 4°C vs 20°C at five different growth phases.   EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 G1 – early lag phase LMON_0762 Hypothetical protein  + lmo0758 20 3.53 -0.35 0.48 -1.47 1.75 LMON_0763 glyoxalase family protein  + lmo0759 20 3.55 0.04 1.28 -0.61 1.78 LMON_0765 Nitrilotriacetate monooxygenase component B  + lmo0761 20 3.66 0.53 1.49 0.70 1.99 LMON_0920 Succinate-semialdehyde dehydrogenase [NAD]  + lmo0913 10 4.64 5.74 2.96 0.40 2.35 LMON_1761 Hypothetical protein  + lmo1694 10 3.78 5.06 2.55 -0.95 0.67 LMON_1936 Pyruvate,phosphate dikinase  + lmo1867 20 4.31 2.17 1.90 0.47 2.16 LMON_2054 Dihydroxy-acid dehydratase ilvD + lmo1983 4* 3.52 3.89 3.34 4.29 3.62 LMON_2055 Acetolactate synthase large subunit ilvB + lmo1984 3/15 3.67 4.17 3.41 4.41 4.37 LMON_2056 Acetolactate synthase small subunit ilvH + lmo1985 15/1 3.67 4.02 3.39 4.41 4.03 LMON_2234 Hypothetical protein   lmo2158 12 3.80 5.25 3.18 0.77 3.03 G2 – transition to exponential growth phase LMON_0020 Hypothetical protein   lmo0019 10 2.62 5.25 3.85 0.65 1.64 LMON_0920 Succinate-semialdehyde dehydrogenase [NAD]  + lmo0913 10 4.64 5.74 2.96 0.40 2.35 LMON_1003 Hypothetical protein   lmo0994 12 3.09 5.04 2.48 -0.56 2.59 LMON_1488 L-carnitine/choline ABC transporter, permease protein  opuCD  lmo1425 10 2.34 5.37 3.52 1.79 1.29 LMON_1489 L-carnitine/choline ABC transporter, substrate-binding protein  opuCC  lmo1426 10 2.45 5.23 3.78 1.82 1.24 LMON_1490 L-carnitine/choline ABC transporter, permease protein  opuCB  lmo1427 10 2.51 5.28 3.78 1.58 1.10 LMON_1491 L-carnitine/choline ABC transporter, ATP-binding protein  opuCA  lmo1428 10 2.63 5.54 3.94 1.30 -0.01 LMON_1761 Hypothetical protein  + lmo1694 10 3.78 5.06 2.55 -0.95 0.67 LMON_2234 Hypothetical protein   lmo2158 12 3.80 5.25 3.18 0.77 3.03 LMON_2306 Arsenate reductase  + lmo2230 10 3.36 4.77 2.95 -1.12 1.49 G3 – mid-exponential growth phase LMON_0369 Twin-arginine translocation protein  tatC  lmo0361 13 -0.08 0.57 4.35 0.25 3.49 LMON_0370 Twin-arginine translocation protein  tatA  lmo0362 13 -0.06 0.94 5.30 0.65 3.26 LMON_0373 Ferrous iron transport permease  efeU + lmo0365 13 -0.66 0.87 4.80 2.03 3.72  166  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_0374 Ferrous iron transport periplasmic protein efeO + lmo0366 13 -1.18 0.87 4.60 1.12 3.74 LMON_0375 Ferrous iron transport peroxidase  efeB + lmo0367 13 -1.08 1.04 4.55 1.33 3.74 LMON_0541 ABC transporter, substrate-binding protein   lmo0541 13 -0.95 -0.08 4.34 0.95 2.00 LMON_0944 Hypothetical protein    10 2.77 3.84 4.22 1.39 2.00 LMON_1016 Hypothetical protein  + lmo1007 13 0.17 1.14 4.89 1.41 3.19 LMON_2261 Cell surface protein IsdA, transfers heme from hemoglobin to apo-IsdC isdA  lmo2185 13 -1.77 -0.61 4.21 0.36 3.24 LMON_2262 NPQTN cell wall anchored protein isdC  lmo2186 13 -1.50 -0.56 4.38 0.95 4.24 G4 – transition to stationary phase LMON_1312 Membrane protein   lmo1252 19/15 -0.07 0.15 -0.01 4.16 1.28 LMON_2054 Dihydroxy-acid dehydratase ilvD + lmo1983 4* 3.52 3.89 3.34 4.29 3.62 LMON_2055 Acetolactate synthase large subunit ilvB + lmo1984 3/15 3.67 4.17 3.41 4.41 4.37 LMON_2056 Acetolactate synthase small subunit ilvH + lmo1985 15/1 3.67 4.02 3.39 4.41 4.03 LMON_2057 Ketol-acid reductoisomerase ilvC + lmo1986 4 2.98 4.08 3.00 3.96 3.28 LMON_2059 3-isopropylmalate dehydrogenase leuB + lmo1988 3 3.18 3.60 2.86 3.84 4.72 LMON_2061 3-isopropylmalate dehydratase small subunit leuD + lmo1990 3 3.16 3.64 2.71 3.96 5.70 LMON_2062 Threonine dehydratase ilvA + lmo1991 3 3.16 3.65 2.50 3.86 5.58 LMON_2286 Hypothetical protein   lmo2210 19 1.59 0.71 2.28 3.79 2.33 LMON_2534 Cell wall-binding protein   lmo2522 4 -1.33 2.20 0.95 4.50 1.70 G5 – late-stationary phase LMON_0740 D-allulose-6-phosphate 3-epimerase   + lmo0735 2 -0.81 -1.52 -2.11 -1.19 10.05 LMON_0741 D-allose-6-phosphate isomerase  + lmo0736 2 -0.52 -1.36 -1.52 -0.35 10.41 LMON_0742 Hypothetical protein  + lmo0737 2 0.04 -0.98 -1.86 1.08 10.49 LMON_0743 PTS system, D-allose-specific   + lmo0738 2 -0.61 -0.29 -0.59 1.52 9.56 LMON_0744 6-phospho-beta-glucosidase  + lmo0739 2 -0.85 -0.15 -0.41 0.86 8.78 LMON_2673 PTS system, lactose/cellobiose specific IIB subunit   lmo2650 2 -2.19 -3.61 -1.53 -1.74 8.51 LMON_2674 PTS system, IIA component   lmo2651 2 -2.41 -4.12 -1.54 -1.80 9.19  167  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_2706 PTS system, cellobiose-specific IIB component  + lmo2683 17 -1.46 -1.30 1.32 -2.04 10.66 LMON_2707 PTS system, cellobiose-specific IIC component  + lmo2684 17 -1.57 -1.40 1.44 -2.83 9.36 LMON_2708 PTS system, beta-glucoside/cellobiose-specific IIA component  + lmo2685 17 -1.38 -1.28 1.33 -2.43 9.28 * indicates <0.50 cluster membership; G1-G5 refer to differential expression in L. monocytogenes cells grown at 4°C relative to 20°C and across five specific growth phases (see Fig 4-1); Blue shading indicates genes with significantly increased gene expression (> 1 log2, p<0.05) and yellow shading indicates genes with significantly decreased expression (< -1 log2, p<0.05) at 4°C relative to 20°C; Bolded values indicate significant (p<0.05) differential expression changes. Genes shaded in yellow are highly expressed at more than one growth phase. 4.3.4.2 Transition to exponential growth phase (G2) At G2, cells have had an adequate amount of time to repair cold-induced damage to molecular structures, and to reprogram their cellular machinery to support growth at low temperatures. As G2 shares characteristics with both the lag and exponential phases, only 22 genes were uniquely upregulated at G2, and only 21 genes were uniquely downregulated. The exclusively induced genes encoded the iron binding protein, Fri; a cold-shock protein, CspD; and a DNA repair and SOS response protein, RecA. Most of these genes exhibited slightly less than a 2-fold increase in expression at G1, maximum increased expression at G2, and no change in expression or decreased expression from G3 - G5, as is characteristic of cluster 10 (Fig 4-5). The upregulation of fri, cspD, and recA have been previously reported in L. monocytogenes in response to cold stress (Bayles and Wilkinson, 2000; Chan et al., 2007b; Hebraud and Guzzo, 2000; Liu et al., 2002). Among the genes upregulated at G2, there was a significant (p<0.05) overrepresentation of those involved in glycerol metabolism and the synthesis of arginine, histidine, and BCAAs  168  (Table 4-3). Among the most induced genes (>36 fold) were those encoding the OpuC carnitine uptake system (opuCABCD), hypothetical proteins, and succinate-semialdehyde dehydrogenase (LMON_0920) which was also strongly induced at G1 (Table 4-5).  4.3.4.3 Mid-exponential growth phase (G3) A total of 47 genes were uniquely induced during exponential growth at 4°C, with most genes exhibiting the DE pattern shown in cluster 13 (Fig 4-5). Aside from a few genes with roles in iron transport (fhuC, fhuB, LMON_2440-2441) and oxidoreductase activity (LMON_0560, LMON_2031, pdhD), most are not known to share any common functions. Among G3-upregulated genes, there was an overrepresentation (p<0.05) of genes involved in glycerol metabolism, aerobic respiration, iron transport, and the biosynthesis of arginine, histidine, and BCAAs (Table 4-3). Other upregulated genes included those involved in the pentose phosphate pathway (LMON_0349-0352, LMON_0647), osmolyte uptake (opuCABCD), and 12 genes fell within PTS operons that facilitate the uptake of mannose, cellobiose, and beta-glucoside. Genes highly induced at G3 (>16-fold) were associated with iron (efeUOB) and heme transport (isdCAEF), as well as protein export (tatAC) (Table 4-5). As expected, many of the proteins previously induced during cold acclimatization (G1 - G2) were downregulated during exponential growth. 4.3.4.4 Transition to stationary phase (G4)  Upon entry into stationary phase, cold-adapted cells downregulated almost 3× more genes than they upregulated (Fig 4-3). Unique to G4 was the upregulation of 53 genes with roles in glycine betaine transport (gbuAB), tRNA processing (LMON_T39, T60, T66), nucleotide biosynthesis (pyrP, pyrB, pyrAa, pyrR, LMON_1838), and maintaining a rod cell shape (mreBD), among others. Most of these genes displayed DE patterns that belonged to clusters 8 or 19 (Fig 4- 169  5), in which DE peaks at G4 and falls dramatically at G5. The five most strongly induced genes (>16-fold) at this growth phase encoded three isoleucine biosynthesis proteins (ilvDBH) and two cell wall-associated proteins (LMON_2534, LMON_1312) (Table 4-5). G4 and G5 also shared a set of exclusively upregulated genes (n=23) with functions in nucleotide biosynthesis (purAHNFE, pyrDII, carB, pyrC), tRNA processing (LMON_T51), and cell-wall recycling (LMON_0029, LMON_2776), among others. All genes exhibited patterns of DE that belonged to cluster 1 or 15 (Fig 4-5). Nucleotide biosynthesis genes are likely activated at these growth phases to accommodate the increased demand for ribonucleic acids imposed by the large abundance of strongly upregulated genes at G5.  4.3.4.5 Late-stationary phase (G5)  Cold-adapted stationary-phase cells exhibited both the largest number of DE expressed genes and the largest magnitude of expression changes (Fig 4-3). Similar results were noted by Chan et al. (2007b) in their L. monocytogenes microarray study, that compared the cold-stress regulons of exponential and stationary-phase cells. More than 50% of G5-upregulated genes exhibited >4-fold increased expression and 17 genes, predominantly associated with PTS mediated sugar uptake systems, were upregulated between 100- and >1,000-fold. Many of the genes strongly induced by cold stress at this growth phase, also exhibited some of the highest expression levels in stationary-phase cells at 20°C, highlighting their overall importance during this growth phase and their enhanced importance in cold-adapted stationary-phase cells. Additionally, 50 - 60% of G5 genes were exclusively DE at this growth phase. The exclusively induced genes had roles in tryptophan (trpBCDFG, aroE) and ATP synthesis (atpACDFGH, adk, LMON_0091), creatine and rhamnose degradation, and ribosome and phage-related processes. Other genes encoded motility- 170  associated proteins (motB, cheV, flaA), an osmosensitive K+ channel histidine kinase (kdpCDE), an Na(+) H(+) antiporter (LMON_2393-2396), the transcription repressor CodY, and low temperature requirement protein B (ltrB). Many genes associated with transcription and translation were also strongly induced at G5 (polC, dnaC, dnaI, topB, rbfA, mfd), including 28 transcription regulators, and 14 genes involved in tRNA processing (Supplementary Table 4-1). Biological processes enriched (p<0.05) among the genes induced at G5 included arginine, histidine, ribonucleotide, and BCAA biosynthesis genes, as well as genes associated with carbohydrate uptake, the pentose phosphate pathway, glycerol and peptidoglycan degradation, and aerobic respiration (Table 4-3). The DE patterns shown in clusters 1, 2, 3, 5, 9, 11, 12, 15, 17, and 20 all describe genes with increased expression in cold-adapted stationary-phase cells, while clusters 4, 6 - 8, 10, 14, 16, 18, and 19 describe genes with limited or no roles in stationary-phase viability at low temperatures (Fig 4-5).  4.3.5 L. monocytogenes cold stress regulon: A comparison with previous studies Numerous previous studies have elucidated a large pool of genes with putative or known roles in the L. monocytogenes CSR (Chan et al., 2008, 2007ab; Cordero et al., 2016; Durack et al., 2013; Liu et al., 2002). In this section, we will determine whether these genes are expressed in a cold-tolerant strain, whether a significant induction occurs when the downshift is from 20°C to 4°C (rather than from 37°C, as in most prior studies), and whether any growth-phase dependencies exist.    171  4.3.5.1 Osmolyte and oligopeptide uptake  In both L. monocytogenes and B. subtilis, low temperature and high osmolarity stress induce the intracellular accumulation of solutes and short peptides. These solutes and peptides function as osmoprotectants that facilitate cell growth under stressful conditions (Angelidis et al., 2002; Angelidis and Smith, 2003; Bayles and Wilkinson, 2000; Beumer et al., 1994; Ko et al., 1994; Ko and Smith, 1999; Wemekamp-Kamphuis et al., 2004a). In L. monocytogenes, carnitine and glycine betaine are the predominant solutes that accumulate (Beumer et al., 1994; Gerhardt et al., 2000; Ko et al., 1994), and their transport systems are encoded by the opuCABCD and gbuABC operons, respectively. Chan et al. (2007b) reported induction of opuCABCD and gbuC in cold-adapted L. monocytogenes exponential- but not stationary-phase cells, while Durack et al., (2013) did not find that either system was induced in late exponential early stationary-phase cells. In the present study, opuCBCD was upregulated in all growth phases, while opuCA was upregulated at G1 - G4 (Table 4-6). Regardless, all genes in the OpuC operon belonged to the DE pattern shown in cluster 10 (Fig 4-5), exhibiting the highest level of induction at G2 (up to 48-fold) and the lowest level at G5, which agrees with previous work (Chan et al., 2007b; Durack et al., 2013). Similar to the findings of Chan et al. (2007b) gbuA and gbuB were only upregulated at G4 (Table 4-6). Oligopeptide uptake in L. monocytogenes is mediated by the membrane permease OppA (oppABCDF) which transports peptides containing up to eight residues (Borezee et al., 2000). This transporter appears to contribute to L. monocytogenes cold tolerance as an oppA null mutant displayed reduced growth at 5°C in BHIB (Borezee et al., 2000). Previously, Durack et al. (2013) observed upregulation of oppBCF in late exponential early stationary-phase cells at 4°C while Chan et al. (2007b) found that only oppA levels were elevated in stationary-phase cells. Contrary to these reports, we did not observe any notable DE of the opp operon, except for a 2.6-fold  172  induction of oppF at G5 (Table 4-6). While the operon may be necessary for low temperature growth, it does not appear to have increased transcription at 4°C relative to 20°C in our cold tolerant strain.  Table 4-6. Genes commonly associated with the L. monocytogenes CSR.  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 Osmolyte and oligopeptide uptake LMON_1024 Glycine betaine ABC transport system, ATP-binding protein gbuA + lmo1014 8 0.15 0.56 0.82 1.68 -0.69 LMON_1025 Glycine betaine ABC transport system, permease protein gbuB + lmo1015 19 0.04 0.48 0.75 1.14 0.08 LMON_1026 Glycine betaine ABC transport system, glycine betaine-binding protein gbuC + lmo1016 15 0.10 0.54 0.64 0.87 0.71 LMON_1488 Osmotically activated L-carnitine/choline ABC transporter, permease protein OpuCD opuCD  lmo1425 10 2.34 5.37 3.52 1.79 1.29 LMON_1489 Osmotically activated L-carnitine/choline ABC transporter, substrate-binding protein OpuCC opuCC  lmo1426 10 2.45 5.23 3.78 1.82 1.24 LMON_1490 Osmotically activated L-carnitine/choline ABC transporter, permease protein OpuCB opuCB  lmo1427 10 2.51 5.28 3.78 1.58 1.10 LMON_1491 Osmotically activated L-carnitine/choline ABC transporter, ATP-binding protein OpuCA opuCA  lmo1428 10 2.63 5.54 3.94 1.30 -0.01 LMON_2268 Oligopeptide transport ATP-binding protein OppF oppF  lmo2192 2 -0.56 -0.61 -0.43 -0.39 1.39 LMON_2269 Oligopeptide transport ATP-binding protein OppD oppD  lmo2193 2 -0.59 -0.74 -0.38 -0.30 0.91 LMON_2270 Oligopeptide transport system permease protein OppC oppC  lmo2194 15* -0.66 -0.91 -0.35 -0.15 -0.29 LMON_2271 Oligopeptide transport system permease protein OppB oppB  lmo2195 19 -0.61 -0.85 -0.35 0.09 -0.79 LMON_2272 Oligopeptide ABC transporter, periplasmic oligopeptide-binding protein OppA oppA  lmo2196 9 -0.67 0.00 -0.46 -1.44 -0.17 RNA and DNA repair LMON_0004 Hypothetical protein  + lmo0004 16 1.25 0.69 0.54 -0.29 -0.69 LMON_0005 DNA recombination and repair protein RecF recF + lmo0005 16 1.40 0.74 0.59 0.06 -0.15 LMON_0006 DNA gyrase subunit B gyrB + lmo0006 20 1.24 0.53 0.49 -0.15 0.45 LMON_0007 DNA gyrase subunit A gyrA + lmo0007 20 1.17 0.61 0.40 -0.10 1.15  173  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_0846 Endoribonuclease L-PSP  + lmo0844 7/6/10 0.53 1.18 1.11 0.24 -0.41 LMON_0848 Excinuclease ABC subunit C  + lmo0846 15 -0.44 -0.04 0.15 1.76 1.12 LMON_0869 Cold-shock DEAD-box protein A  + lmo0866 14 1.97 0.85 0.38 1.87 0.45 LMON_1037 Ribonuclease J1 (endonuclease and 5' exonuclease)   lmo1027 16 1.71 0.33 -0.03 -0.10 -1.17 LMON_1225 DNA polymerase X family  + lmo1231 2 1.12 0.22 0.42 0.76 2.10 LMON_1305 ATP-dependent RNA helicase YxiN  + lmo1246 6 0.99 0.75 0.56 0.07 -1.04 LMON_1317 Hypothetical protein  + lmo1256 12 2.04 1.68 1.00 -1.20 2.14 LMON_1336 DNA topoisomerase I topA + lmo1275 4 0.02 0.74 0.30 1.00 0.31 LMON_1348 Topoisomerase IV subunit A parC + lmo1287 2 -0.02 -0.06 0.19 0.10 1.39 LMON_1461 RecA protein recA + lmo1398 12 0.73 1.27 0.34 -0.27 0.46 LMON_1513 ATP-dependent RNA helicase YqfR   lmo1450 6 1.33 1.12 1.18 1.01 -1.04 LMON_1787 ATP-dependent RNA helicase YfmL   lmo1722 6 1.69 1.45 1.28 1.21 0.47 LMON_2342 ATP-dependent nuclease, subunit A   lmo2267 17 -0.99 -0.77 -0.35 -0.62 1.77 LMON_2343 ATP-dependent nuclease, subunit B addB  lmo2268 17 -0.78 -0.45 -0.11 -0.60 1.42 LMON_2778 DNA topoisomerase III topB  lmo2756 17 -0.25 -0.03 0.19 -0.20 1.40 LMON_2812 DNA-binding protein   lmo2792 5 0.62 1.34 0.75 -0.80 2.49 Regulatory elements LMON_0198 Virulence regulatory factor PrfA prfA  lmo0200 17 -2.04 -0.44 1.01 0.10 5.32 LMON_0229 Transcriptional regulator CtsR ctsR + lmo0229 16 1.90 0.42 0.47 -0.55 0.16 LMON_0244 RNA polymerase sigma factor SigH sigH + lmo0243 7 -1.29 -0.60 -0.49 -0.59 -2.35 LMON_0431 RNA polymerase sigma factor SigC sigC  lmo0423 13 -2.03 -1.48 -0.59 -0.62 -0.97 LMON_0679 Motility gene repressor MogR  mogR  lmo0674 16 0.62 0.02 0.03 -1.20 -0.72 LMON_0900 RNA polymerase sigma factor SigB sigB + lmo0895 10 0.00 0.58 0.19 -0.78 -0.62 LMON_1166 Ethanolamine sensory transduction histidine kinase  + lmo1173 5/2 1.03 0.71 -0.03 0.06 2.43 LMON_1341 GTP-sensing transcriptional pleiotropic repressor codY codY + lmo1280 3 -0.42 0.61 0.02 0.09 1.27 LMON_1517 RNA polymerase sigma factor RpoD rpoD  lmo1454 5 0.08 -0.24 -0.60 -0.88 0.54 LMON_1539 Heat-inducible transcription repressor HrcA hrcA  lmo1475 16 1.42 -0.02 0.10 -0.28 -0.43 LMON_1807 Two-component sensor histidine kinase BceS cesK  lmo1741 2 -0.17 -0.17 0.02 0.03 2.05 LMON_1811 Two-component response regulator YvcP (VirR in L. monocytogenes 10403S) virR  lmo1745 17 -0.48 -0.61 -0.12 -0.30 0.02 LMON_2432 Sensor histidine kinase virS  lmo2421 14 1.27 -0.03 0.28 0.89 0.17 LMON_2472 RNA polymerase sigma factor SigL sigL  lmo2461 5 -0.20 -0.05 -0.10 -0.42 0.85 LMON_2527 Transcriptional regulator DegU degU  lmo2515 16 1.61 0.62 0.54 0.04 -0.94  174  EGD ORF EGD description Gene abbr. Strand EGD-e ORF Cluster memb-ership Log2 DE across five growth phases G1 G2 G3 G4 G5 LMON_2702 Osmosensitive K+ channel histidine kinase KdpD kdpD  lmo2679 2 0.99 0.46 0.60 0.44 4.22 Ribosome functions LMON_0209 LSU ribosomal protein L25p ctc + lmo0211 10 2.16 3.43 2.72 0.87 1.58 LMON_0486 LSU ribosomal protein L32p rpmF + lmo0486 19 1.43 1.46 1.76 1.75 1.22 LMON_1388 Translation initiation factor 2 infB + lmo1325 2 0.55 0.09 -0.01 0.39 1.16 LMON_1390 Ribosome-binding factor A rbfA + lmo1327 2 0.59 0.16 -0.03 0.25 1.75 LMON_2523 Ribosomal subunit interface protein   lmo2511 5 1.08 1.32 0.65 -0.90 3.14 Cold-stress proteins LMON_1427 Cold-shock protein cspL + lmo1364 16 1.58 1.34 0.45 0.12 0.27 LMON_1947 Cold shock protein CspD cspD + lmo1879 12 0.96 1.00 -0.09 -3.93 -1.96 LMON_2087 Cold shock protein CspB cspB  lmo2016 16 -3.67 -4.42 -5.85 -7.83 -2.97 Additional proteins LMON_0213 Low temperature requirement B protein ltrB + lmo0215 2 0.03 -0.26 -0.25 0.00 1.78 LMON_0398 Low temperature requirement protein A ltrA  lmo0389 12* 0.43 -0.23 0.61 -0.63 0.08 LMON_0691 Flagellar motor rotation protein MotB motB + lmo0686 17 -0.81 -0.37 0.07 -0.26 1.44 LMON_0693 Glycosyl transferase gmaR + lmo0688 17 -0.84 -0.42 -0.06 -0.44 1.29 LMON_0694 Chemotaxis protein CheV cheV + lmo0689 17 -0.77 -0.04 -0.19 -0.65 2.37 LMON_0695 Flagellin protein FlaA flaA + lmo0690 5 0.58 -1.03 -0.70 -2.00 1.16 LMON_0950 Non-specific DNA-binding protein Dps / Iron-binding ferritin-like antioxidant protein / Ferroxidase fri + lmo0943 10 0.39 1.50 0.03 -0.51 -0.22 LMON_2409 Low temperature requirement C protein ltrC + lmo2398 12 1.14 1.21 0.75 -1.61 0.16 * indicates <0.50 cluster membership; G1-G5 refer to differential expression in L. monocytogenes cells grown at 4°C relative to 20°C and across five specific growth phases (see Fig 4-1); Blue shading indicates genes with significantly increased (> 1log2, p<0.05) gene expression and yellow shading indicates genes with significantly decreased (< -1log2, p<0.05) expression at 4°C relative to 20°C; Bolded values indicate significant (p<0.05) differential expression changes.  4.3.5.2 RNA and DNA repair  At low temperatures, both DNA replication and transcription are hindered by cold-induced changes in nucleic acid structures. L. monocytogenes appears to respond to these challenges by upregulating genes encoding topoisomerases and DNA gyrases, which help maintain the superhelical tension of DNA; RNA helicases, which unwind secondary RNA structures; and exo- and endonucleases, which function in nucleic acid repair (Chan et al., 2007b; Durack et al., 2013; Markkula et al., 2012). Markkula et al. (2012) reported the upregulation of four DEAD-box RNA  175  helicase genes (lmo0866, lmo1246, lmo1450, lmo1722) in L. monocytogenes for up to 7 h following a downshift from 37°C to 5°C. They also found that, compared with the wildtype strain, null mutants of lmo0866, lmo1450, and lmo1722 displayed restricted growth at 3°C. Similarly, we observed upregulation of lmo1450 (LMON_1513) and lmo1722 (LMON_1787) at G1 - G4, and lmo0866 (LMON_0869) at G1:G4 (Table 4-6). Also in agreement with Markkula et al. (2012), lmo1246 (LMON_1305) appeared to be the least important of the four genes, with <2-fold induced expression at G1 - G3.     Several DNA gyrase and topoisomerase genes were also upregulated in our study. Notably, topA was upregulated at G4, and topB and parC were upregulated at G5 (Table 4-6). In L. monocytogenes, the DNA gyrase genes gyrB and gyrA are in a four-gene operon, along with a gene encoding a hypothetical protein, and recF, a DNA recombination and repair protein. All four genes were upregulated at G1, and gyrA was also induced at G5. Upregulation of another DNA recombination and repair protein, recA, was also observed at G2.  Among the nucleic acid repair-proteins induced in this study were ribonuclease J1 (LMON_1037) at G1, an endoribonuclease (LMON_0846) at G2 - G3, two nuclease subunits (LMON_2342-2343) at G5, and an excinuclease (LMON_0848) at G4 - G5 (Table 4-6). Additionally, an X family DNA polymerase (LMON_1225) was induced at G1:G5. These polymerases synthesize unusual DNA structures that may serve as indicators of the induction of DNA repair (Mahajan et al., 2002; Ramadan et al., 2004; Showalter et al., 2001). A DNA-binding protein (LMON_2812) with a putative role in oxidative-damage protection (Nair and Finkel, 2004) was also upregulated at G2 and G5. Lastly, a hypothetical protein (LMON_1317) containing a nudix hydrolase domain was upregulated at G1 - G3:G5 (Table 4-6). Some members of this protein family are known to degrade oxidatively damaged nucleoside di- and triphosphates, while other  176  members control the levels of metabolic intermediates and signaling compounds (McLennan, 2006).  Overall, genes associated with RNA and DNA repair were predominantly cold-induced in lag- and stationary-phase cells (G1 and G5). This is expected given that cold-induced damage to nucleic acids is highest directly following cold stress, and nutrient-depleted stationary-phase cultures experience higher mutation rates and increased levels of oxidative damage (Ballesteros et al., 2001; Díaz-Acosta et al., 2006; Fredriksson and Nyström, 2006; Navarro Llorens et al., 2010).  4.3.5.3 Regulatory elements  Many mutant characterization and transcriptome studies have evaluated the roles of alternative sigma factors (σB, σC, σH, σL), two-component regulatory systems (TCRSs), and negative regulators in the L. monocytogenes CSR. Of the four alternative sigma factors, σB has been the most extensively studied. It positively regulates over 100 genes when the organism enters stationary phase or is subjected to environmental stresses including low pH, high salt, or carbon starvation (Becker et al., 1998; Ferreira et al., 2001; Fraser et al., 2003; Kazmierczak et al., 2003; Wiedmann et al., 1998). Although induced expression of sigB has been observed in L. monocytogenes up to 12 h following cold stress (Arguedas-Villa et al., 2010; Becker et al., 2000; Chan et al., 2007b), the σB regulon is predominantly downregulated during cold stress (Chan et al., 2007b; Durack et al., 2013), demonstrating that increased sigB expression does not necessarily correlate with increased σB activity. Furthermore, compared with wildtype strains, sigB null mutants show little to no difference in cold tolerance (Becker et al., 2000; Chan et al., 2008, 2007a; Moorhead and Dykes, 2004), indicating that σB is not critical for cold-stress survival. In the present study, the DE pattern of sigB followed that of cluster 10 (Fig 4-5), with a maximum 1.5-fold  177  increase in expression at G2 (Table 4-6). Similar expression patterns were observed for three additional σB operon genes (rsbWVX) while the first four genes (rsbRSTU) of the eight-gene operon were not induced under cold stress. Despite low levels of sigB induction, we found that σB-dependent genes were overrepresented (p<0.05) among the cold-induced genes at all growth phases (Table 4-7). Examples of such genes include fri, bsh, dapE, uspA, gabD, opuCABCD, mpoBACD, phoU, csbD, ltrC, and hrcA. Although some of these genes are solely activated by σB, others are co-regulated by other unknown or known transcription factors such as σL, σH, and PrfA (Chan et al., 2008, 2007a; Chaturongakul et al., 2011; Dussurget et al., 2002; Polidoro et al., 2002). As previously mentioned, σB is also activated upon entry into stationary phase (Boylan et al., 1993; DeMaio et al., 1996; Kullik and Giachino, 1997). Consistent with this fact, we observed that sigB transcript levels during growth at 20°C increased from 4.5k mapped reads at C1 to 15k at C5 (Supplementary Table 4-2). The alternative sigma factor σL has been shown to contribute to the ability of L. monocytogenes to tolerate cold (Chan et al., 2008; Raimann et al., 2009), osmotic, and acid stress (Okada et al., 2006; Raimann et al., 2009). This sigma factor positively regulates >400 genes, including those involved in cell envelope synthesis, motility, and PTS sugar uptake and catabolism (Arous et al., 2004; Mattila et al., 2012). Although sigL induction has been reported in exponential and late exponential early stationary-phase cells of cold-adapted L. monocytogenes (Durack et al., 2013; Liu et al., 2002; Raimann et al., 2009), deleting sigL does not impact the ability of L. monocytogenes ability to grow at cold temperatures (Chan et al., 2008; Mattila et al., 2012). In the present study, the DE pattern of sigL followed that of cluster 5 (Fig 4-5), with baseline expression at G1 - G4 and <2-fold increased expression at G5 (Table 4-6). Sixteen genes previously identified  178  as being positively regulated by σL in L. monocytogenes at 3°C (Mattila et al., 2012) also exhibited  similar DE patterns.  The remaining alternative sigma factors, σH and σC, contribute to the survival of L. monocytogenes under alkaline and heat stress, respectively (Phan‐Thanh and Mahouin, 1999; Rea et al., 2004; Zhang et al., 2005). However, these sigma factors have limited roles in cold tolerance (Chan et al., 2008). Consistent with previous findings, we did not observe the upregulation of sigH or sigC in our study.  RpoD is the principal RNA polymerase sigma factor in L. monocytogenes and regulates housekeeping genes associated with ribosome structure, protein synthesis, and rRNA and tRNA (Metzger et al., 1994). Like the DE pattern for sigL, the DE pattern of rpoD followed that of cluster 5 (Fig 4-5), with a maximum 1.5-fold increase in expression at G5 (Table 4-6). Accordingly, genes positively regulated by RpoD were overrepresented (p<0.05) among the downregulated genes at all growth phases except G4. Many genes in the RpoD regulon are co-activated by PrfA, and correspondingly, PrfA-regulated genes were also overrepresented among the downregulated genes at G1 - G3:G5 (Table 4-7). Genes co-regulated by RpoD and PrfA include prfA, plcAB, hly, mpl, inlC, and actA, all of which are associated with virulence (Farber and Peterkin, 1991). Other studies have also reported lower levels of PrfA-regulated virulence genes in cold-adapted exponential and stationary-phase cells (Chan et al., 2007b; Durack et al., 2013; Ivy et al., 2012). In the present study, the DE pattern of prfA fit that of cluster 17 (Fig 4-5), with 2-fold and 40-fold increased expression at G3 and G5, respectively (Table 4-6). Interestingly, although the transcription of prfA increased dramatically at G5, the expression levels of many PrfA-dependent virulence genes remained strongly downregulated. Researchers have proposed that PrfA activity, but not prfA transcription, is inhibited by unphosphorylated forms of PTS permeases which occur  179  during active sugar transport; these PTS permeases may bind directly to PrfA (Aké et al., 2011; Mertins et al., 2007; Stoll et al., 2008). Our results support this hypothesis, as many PTS operons and other carbohydrate uptake and catabolism genes were upregulated at G5 and belonged to the same cluster profile as prfA (Fig 4-5, cluster 17). Whether PrfA is subsequently involved in regulating these genes is still unclear.  CodY is a pleiotropic transcriptional regulator that actively represses the transcription of genes involved in amino acid metabolism, nitrogen assimilation, mobility and chemotaxis, and sugar uptake, among others (Bennett et al., 2007). In agreement with previous work (Chan et al., 2007b), we observed the highest induction (2.4-fold) of codY in at G5 following the DE pattern shown in cluster 3 (Fig 4-5).  DegU is another well-known response regulator that regulates the expression of motility-, virulence-, and biofilm-related genes in L. monocytogenes (Gueriri et al., 2008; Knudsen et al., 2004; Williams et al., 2005). Our results showed that degU induction was greatest (3-fold) at G1 (Table 4-6), with a DE pattern consistent with that of cluster 16 (Fig 4-5). This probably explains why increased expression of degU was not observed in previous cold stress transcriptome studies of L. monocytogenes, that focused on exponential- and stationary-phase cells (Chan et al., 2007b; Durack et al., 2013). Despite the induction of degU at G1, the flagella operon (LMON_0680-0694) that the DegU protein regulates remained suppressed, likely due to co-induction of the gene encoding the transcriptional repressor MogR, which shared the same DE pattern as degU. In B. subtilis, DegU is part of the two-component system DegS/U, which regulates the expression of genes encoding various extracellular enzymes (Antelmann et al., 2001). This suggests that DegU may also contribute to the regulation of extracellular enzymes in L. monocytogenes that facilitate cold growth.   180  The negative regulators HrcA and CtsR repress expression of several genes for heat-shock proteins and cellular protein quality control (e.g. dnaK, grpE, groES, groEL, clpB, clpC, clpE, clpP) in L. monocytogenes and other bacteria (Hanawa et al., 2000; Karatzas et al., 2003; Nair et al., 2000). Increased expression of many of these genes has also been reported in response to salt, cold, and ethanol stress (Kilstrup et al., 1997; Liu et al., 2002; Salotra et al., 1995). In the present study, we found that hrcA and ctsR, like degU and mogR, were both maximally upregulated at G1, as were many of the genes they regulate. This was somewhat surprising given the inverse relationship expected between the DE patterns of transcription repressors and their gene targets. However, CtsR and HrcA regulons are known to have a considerable amount of overlap with PrfA and σB regulons, demonstrating the complexity and fine-tuning abilities of bacterial regulatory networks, which allow bacteria to survive a wide-range of conditions (Chaturongakul et al., 2011; Hu et al., 2007a, 2007b).  Two-component signaling systems (TCSs) are also important regulators of bacterial stress responses and typically consist of a transmembrane sensor histidine kinase (HK), and a cognate cytoplasmic response regulator (Chang and Stewart, 1998; Krell et al., 2010; Stock et al., 2000; West and Stock, 2001). The sequenced genome of L. monocytogenes EGD-e contains 16 known TCSs (Glaser et al., 2001). Using mutant characterization, Pontinen et al. (2015) showed that only the HKs LisK and YycG of the LisKR and YycGF TCSs, respectively, are important for L. monocytogenes growth under cold stress. However, increased expression of lisK was not observed in either the Pontinen et al. (2015) study or the present study, highlighting that the importance of a gene does not necessarily correlate with its level of induction. HKs with increased expression in our study included cesK and LMON_1166 at G1, and kdpD, LMON_1166, and virS at G5 (Table 4-6). Of these, kdpD, which encodes the HK for an osmosensitive K+ channel, exhibited the largest  181  increase in expression (18.6-fold). KdpD, together with its response regulator KdpE, control the expression of a high-affinity K+ translocating ATPase (Csonka and Epstein, 1996; Jung et al., 1997). Among the many K+ transport systems in bacteria, Kdp has the highest affinity for K+ and it is only expressed when other systems are unable to meet the cell’s K+ needs. Our results therefore suggest that Kdp plays an important role in long-term cold stress survival of L. monocytogenes. Table 4-7. Transcription regulators significantly (p<0.05*) overrepresented among genes differentially expressed at 4°C vs. 20°C.  Regulators Regulon gene examples p-value (# of contributing genes) G1 G2 G3 G4 G5 Regulons upregulated σB inlA, bsh, mpoABCD, gabDE,opuCABCD, ltrC, csbD, phoU, hrcA 3.70E-10 (63) 3.95E-23 (69)  1.26E-2 (8) 9.58E-4 (51) CodY argBDFJ   1.96E-2 (4)  4.70E-2 (4) CtsR clpBCEP, mcsAB, ctsR, gpmA 1.74E-4 (8)     HrcA  hrcA, grpE, dnaK 4.14E-2 (3)     Regulons downregulated RpoD dltABCD, plcAB, hly, mpl, actA, pmk, pgdA, ctaP, arcB 1.91E-2 (24) 4.57E-2 (16) 1.05E-2 (14)  1.59E-2 (39) VirR dltABCD 2.15E-2 (4) 3.50E-4 (5) 1.66E-3 (4)   PrfA prfA, mpl, actA, plcB, inlC, hly 1.16E-2 (5)  4.44E-4 (5)  3.14E-2 (6) σB mogR, mpoAB, phoU, ltrC, arsC, uspA    2.14E-3 (52)  MogR fliNPQRK, cheRY, flhB, flgDE     1.35E-2 (11) * Statistical overrepresentations of gene sets were determined using Fisher’s exact test and the BioCyc database  182  4.3.5.4 Ribosome functions  When bacteria are subjected to a temperature downshift, their ribosome structures become compromised due to thermodynamic induced structural changes to proteins, resulting in translation inhibition and a prolonged lag phase. In fact, inhibition of ribosomal functions induces CSR proteins (Jones and Inouye, 1996). In E. coli, three ribosome-associated proteins (IF2, CsdA, RbfA) are required for protein synthesis and subsequent cell growth at low temperatures (Dammel and Noller, 1995; Gualerzi and Pon, 1990; Toone et al., 1991). In contrast, 22 other ribosomal proteins are not believed to be essential for the growth of E. coli or B. subtilis at low temperatures (Akanuma et al., 2012; Shoji et al., 2011). In L. monocytogenes, the role of ribosomes and their associated proteins in cold stress remains largely unknown. In a study by Durack et al. (2013), ribosome protein genes were strongly activated in osmo- and cold-adapted L. monocytogenes cells. In our study, of the 58 ribosomal proteins identified in L. monocytogenes EGD, 22 were upregulated exclusively at G5 in the cold tolerant- strain we studied. Among these genes were those encoding L. monocytogenes homologs of IF2 (infB) and RbfA, previously mentioned for their roles in stabilizing cold-sensitive ribosomes in E. coli. An additional two ribosomal protein genes, ctc and rpmF, were significantly upregulated at all growth phases, and a ribosomal subunit interface protein (LMON_2523) was induced at G1 - G2:G5 (Table 4-6). In B. subtilis, ctc is induced in response to osmotic, heat, and oxidative stress (Hecker and Völker, 1998; Völker et al., 1994), and it has similarly been linked to osmo- and cold-tolerance in L. monocytogenes (Durack et al., 2013; Gardan et al., 2003).  In bacteria, rRNA and ribosomal protein synthesis are tightly controlled, to meet the translational needs of the cell. The increased transcription of ribosome proteins in cold-adapted stationary-phase cells may reflect an increased demand for protein synthesis, as suggested by the  183  large number of strongly upregulated genes at G5. Alternatively, ribosomal proteins have been shown to participate in extra-ribosomal functions, as independent polypeptides with roles in transcription and DNA repair (Lindström, 2009; Mazumder et al., 2003; Zimmermann, 2003). Thus, they might contribute to the L. monocytogenes CSR in yet undetermined ways. 4.3.5.5 Cold-stress proteins  CSPs are a conserved family of small (~70 aa) proteins containing a nucleic acid-binding domain. Found in many prokaryotic and eukaryotic organisms, CSPs can act as transcriptional activators, antiterminators, or as RNA chaperones that enhance translation at low temperatures by blocking the development of secondary mRNA structures (Hunger et al., 2006; Jiang et al., 1997; Phadtare, 2004). Three CSPs have been identified in L. monocytogenes and are listed here in the order of functional importance: CspL>CspD>CspB (Schmid et al., 2009). Furthermore, L. monocytogenes CSPs appear to be only induced during the early stages following cold stress (Chan et al., 2007b; Durack et al., 2013; Schmid et al., 2009). In the present study, cspB was downregulated at all growth phases, reaching a maximum 228-fold decrease at G5, while cspD and cspL were upregulated at G1 - G2 and then exhibited no change or decreased expression at the later growth phases (Table 4-6). 4.3.5.6 Additional proteins with putative roles in the L. monocytogenes cold-stress response  A few other genes have also been associated with the L. monocytogenes CSR. Zheng and Kathariou (1994, 1995) identified three low temperature requirement proteins (LtrA, LtrB, and LtrC) necessary for growth at cold temperatures. However, varying results have been reported regarding the expression of these genes at low temperatures. Pieta et al. (2014) observed higher  184  levels of ltrC transcripts at 7°C than at 37°C in exponential-phase cells, whereas other researchers saw either no difference in expression or decreased expression in exponential- and stationary-phase cells of cold-adapted L. monocytogenes (Chan et al., 2007b; Durack et al., 2013; Liu et al., 2002). Increased expression of ltrC has also been reported in L. monocytogenes EGD-e, directly following an upshift in temperature from 37°C to 48°C (Olsen et al., 2005). Importantly, we observed increased expression of ltrC only in the early stages following cold stress (G1 and G2), which may partly explain why no differences were seen in studies that analyzed exponential- and stationary-phase cells. As for other low temperature requirement protein genes, ltrB was upregulated at G5 while no notable changes were seen for ltrA.  The fri (flp) gene, which encodes ferritin, is also commonly discussed in cold stress studies of L. monocytogenes and is hypothesized to play a role in iron storage (Ilari et al., 2000; S. J. Olsen et al., 2005; Polidoro et al., 2002). Previous studies have reported that fri is induced upon entry into stationary phase in L. monocytogenes subjected to low temperatures (Hebraud and Guzzo, 2000; Liu et al., 2002; Polidoro et al., 2002). Additionally, fri null mutants display reduced growth at 4°C in BHIB and increased sensitivity to oxidative stress (Dussurget et al., 2005; Olsen et al., 2005). In our study, fri was exclusively induced at G2 (Table 4-6). However, in agreement with the findings of Polidoro et al. (2002), the number of fri transcripts at 20°C was nine times higher in stationary-phase cells than in lag-phase cells. Liu et al. (2006) identified a membrane-associated phosphohydrolase (pgpH) as the interrupted gene in an L. monocytogenes cold-sensitive transposon mutant. Compared with the parent strain, this mutant also showed increased intracellular levels of the phosphorylated guanosine nucleotide (p)ppGpp, suggesting that PgpH may be critical for (p)ppGpp degradation at low temperatures. Cellular (p)ppGpp inhibits RNA synthesis when a shortage of amino acids is  185  present. Such conditions can occur during cold stress, because of structural damage to membrane transporters and intracellular enzymes. Arguedas-Villa et al. (2010) reported increased pgpH expression in cold-tolerant but not cold-sensitive strains of L. monocytogenes at 4°C compared to 37°C. Though we also evaluated the gene expression of a cold-tolerant strain, we observed no significant induction of pgpH (LMON_1529) at 4°C relative to 20°C.  The induction of flagella biosynthesis and motility-related genes has frequently been reported in L. monocytogenes following a temperature downshift (Chan et al., 2007b; Liu et al., 2002; Pieta et al., 2014). However, the reliability of these observations are debatable as 37°C was uniformly used as the control temperature and L. monocytogenes is not typically motile above 30°C (Peel et al., 1988; Way et al., 2004). Recently, Cordero et al. (2016) reported that fast cold-growing L. monocytogenes strains are less motile than slow cold-growing strains. They hypothesized that low motility might allow cold-tolerant strains to proliferate more rapidly at low temperature. In the present study, our cold-tolerant strain exhibited decreased expression of most flagella operons at all growth phases. However, five motility-specific genes (LMON_0691-0695) were upregulated at G5, including motB, gmaR, cheV, and flaA (Table 4-6). The roles of motility-related genes in extended cold-stress survival remain speculative, but one hypothesis is that it is beneficial for cells to be able to move to environments more suitable in terms of nutrition or other requirements. 4.3.6 Cold-induced membrane lipid composition changes To gain a better understanding of both the timing associated with cold-induced membrane FA changes in L. monocytogenes and the types of changes that occur, we analyzed FAs extracted from cells at the same time points used in our RNA-seq experiment.  186  4.3.6.1 Increase in anteiso C15:0 The L. monocytogenes lipid membrane consists predominantly of BCFAs (~90% of total FAs). The four most abundant BCFAs, listed in decreasing order, are anteiso-C15:0 (a-C15:0), a-C17:0, iso-C15:0 and i-C17:0 (Annous et al., 1997; Juneja et al., 1998; Verheul et al., 1997; Zhu et al., 2005a). When L. monocytogenes experiences a decrease in temperature, it increases the relative proportion of a-C15:0 at the expense of a-C17:0 (Kaneda, 1991; Zhu et al., 2005a). Depending on the length of the BCFA and the bacterial strain studied, a switch from iso to anteiso FAs can also occur (e.g., i-C17:0 to a-C17:0).  To date, the membrane FA profile of L. monocytogenes under cold stresses has primarily been investigated using late exponential to early stationary-phase cells (Annous et al., 1997; Juneja et al., 1998; Verheul et al., 1997; Zhu et al., 2005a). In most cases, researchers have observed a ~20% increase in the relative proportion of a-C15:0 among cold-grown cells (5-10°C), with maximum levels ranging from 66 - 80%. Correspondingly, a-C17:0 levels decrease by 20 - 25% with minimum levels ranging from 3 - 14%.  In the present study, a-C15:0 levels increased from 46% at T1 to a maximum of 70% at T4 (Fig 4-6A), whereas levels remained constant around 50% in 20°C-grown cells. Thus, in this study, the first one to look at cold-induced membrane FA changes at multiple growth phases, we show that L. monocytogenes continues to make alterations until it transitions into stationary phase (T4), at which point minimal further adjustments are made. Correspondingly, a-C17:0 levels decreased from 12% at T1 to 4% at T4, whereas at 20°C they ranged from 13 - 18% (Fig 4-6C). Levels of i-C16:0 and iC15:0 also decreased by 2.3% and 9%, respectively at 4°C (Fig 4-6B). Additional changes in the BCFA composition of cold-stressed cells included a 2.5 - 2.9% increase in the proportion of i-C14:0 and the appearance of i-C13:0 and a-C13:0.   187  As previously discussed, anteiso BCFAs are synthesized from the BCAAs isoleucine, and iso BCFAs are synthesized from leucine and valine (Kaneda, 1991). These BCAAs are produced by proteins encoded by a nine-gene operon (LMON_2054-2062). At 4°C, we found that all genes in this operon were induced >4-fold at all growth phases. Interestingly, genes from the BKD operon (LMON_1432-1437) which encodes enzymes that convert BCAAs into BCFAs, showed either no change in expression or decreased expression at 4°C. Other researchers have also noted a lack of bkd induction in L. monocytogenes under cold-stress conditions (Annous et al., 1997; Zhu et al., 2005a). A BKD-independent pathway may exist, in which BCFA synthesis uses branched-chain α-keto acids instead of branched-chain acyl-CoA esters as primers (Kaneda, 1991). Alternatively, Nickel and colleagues (2004) showed that in B. subtilis the promoter regions of five operons, including BKD, are not cold-inducible but exhibited increased mRNA stability at low temperatures.  In addition to the BKD operon, many other FA synthesis genes were also downregulated at 4°C (e.g. fabD, fabG, fabZ, acpP, plsX, plsY). However, a fabG isozyme (LMON_0351) encoding a 3-oxoacyl-[acyl-carrier protein] reductase remained upregulated at G1 - G3:G5. Other upregulated genes with known or putative functions in FA biosynthesis included two short-chain dehydrogenases (LMON_0674 and LMON_1898), an oxidoreductase (ylbE), and a two-component response regulator (LMON_0289) with homology to a membrane FA regulator (YccF) in Streptococcus pneumoniae (Mohedano et al., 2005; Ng et al., 2004).   188   Figure 4-6. Relative proportions of specific branched-chain fatty acids (FAs) out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4°C and 20°C. A) anteiso C15:0, B) iso C15:0, and C) anteiso C17:0. Lipids were extracted from cell concentrates and resulting fatty acid methyl esters (FAME) were analyzed by gas chromatography. See Fig 4-1 for information about growth phases. 4.3.6.2 Shortening of fatty acid chain lengths  Following a temperature downshift, L. monocytogenes incorporates FAs with shorter chain lengths to further decrease the phase-transition temperature of its membrane (Zhu et al., 2005b). Overall, we observed an 8% increase in ≤C14 FAs at 4°C relative to 20°C, which became noticeable at T3 (Fig 4-7A). Moreover, by T5, 92% of Lm1’s membrane FAs contained fewer than 15 carbons; in 20°C grown cells, this percentage was 75%. Our results suggest that L. monocytogenes begins degrading unfavorable membrane phospholipids during lag phase but cannot synthesize its preferred FAs until growth resumes. This poses the question of how the cell retains adequate membrane fluidity during the early stages of cold stress. The answer to this question appears to involve the desaturation of existing FAs.    189  Figure 4-7. Relative proportion of short chain and unsaturated fatty acids (UFAs) out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4 and 20°C. A) Proportion of short chain fatty acids containing ≤14 carbons, B) Proportion of C16:1 cis-Δ9 (palmitoleic acid), C18:1 cis-Δ9 (oleic acid), and all remaining UFAs. Lipids were extracted from cell concentrates and resulting fatty acid methyl esters (FAME) were analyzed by gas chromatography. See Fig 4-1 for information about growth phases. 4.3.6.3 Increase in unsaturated fatty acids  Several bacterial species including E. coli, some species belonging to the phylum cyanobacteria, and some species belonging to the genus Bacillus, increase their proportion of membrane unsaturated FAs (UFAs) directly following a temperature downshift (Aguilar et al., 1998; Sakamoto and Murata, 2002; Weber et al., 2001). This process occurs rapidly due to the presence of membrane-associated desaturases that introduce double bonds into pre-existing saturated FAs (SFAs). For example, in E. coli, cis-vaccenate is produced 30 seconds following a shift from 42°C to 24°C (Garwin and Cronan, 1980).  To date, the roles of UFAs in the L. monocytogenes CSR have been largely overshadowed by interests in BCFA modulation. Although some papers have reported the absence of UFAs in the L. monocytogenes membrane (Mastronicolis et al., 2010, 2005; Miladi et al., 2013; Zhu et al., 2005a), others have detected their presence in small amounts under various conditions (Bisbiroulas  190  et al., 2011; Gianotti et al., 2008; Juneja et al., 1998; Mastronicolis et al., 2010; Raines et al., 1968; Vadyvaloo et al., 2002). In a study of L. innocua, levels of C18:1 FAs increased to 6.2% following a downshift from 35°C to 10°C (Moorman et al., 2008). Similarly, another study found that the Bacillus megaterium membrane contained a maximum of 22% C16:1 cis-Δ9 (palmitoleic acid) at 10°C, whereas the maximum was 2% at 35°C (Suutari and Laakso, 1992). In the present study, the proportions of C16:1 cis-Δ9 and C18:1 cis-Δ9 (oleic acid) peaked at T3 (7%) and T2 (20%), respectively (Fig 4-7B). By contrast, no difference was detected between the proportions of oleic and palmitoleic acid in T4 and C4 cells, or T5 and C5 cells. Very small percentages (<1.2%) of other UFAs were also detected in our samples, but no notable differences were observed between the temperatures treatments (Supplementary Table 4-3). The fact that membrane UFA proportions only increased in cold-stressed lag- and exponential-phase cells likely explains why studies that have investigated late exponential- and stationary-phase cells have failed to detect any changes. Unlike the rapid conversion of SFAs to UFAs, switching from iso to anteiso BCFAs and from longer to shorter acyl chain FAs requires de novo synthesis of whole lipid molecules by cytoplasmic enzymes that are usually linked to growth (Gounot and Russell, 1999). Sato and Murata (1980) showed that following a 10 - 15°C downshift in temperature, cyanobacterial cells could only resume growth and FA biosynthesis once a certain degree of membrane unsaturation was reached. Based on our findings, we suggest that in response to cold stress, L. monocytogenes converts C18:0 and C16:0 to oleic and palmitoleic acid, respectively, to rapidly lower its membrane phase-transition temperature. Fig 4-8 shows how UFA levels appear to compensate for a-C15:0 until its optimal levels are reached at T4. The centre placement of the cis-double bond in oleic and palmitoleic acid makes these FAs highly efficient at increasing membrane fluidity. For example, the average melting points of oleic and palmitoleic acid are 13°C and 1.22°C,  191  respectively, compared to 24°C and 52°C for a-C15:0 and i-C15:0, respectively (Knothe and Dunn, 2009). While oleic and palmitoleic acids are effective at rapidly increasing the membrane fluidity of L. monocytogenes, this membrane profile is probably less stable than one that includes a high proportion of a-C15:0. In support of this idea, Juneja and Davidson (1993) showed that when provided with an external source of oleic acid, L. monocytogenes was able to increase the relative proportion of this acid in its membrane from 0.72 to 28%. However, these cells were highly susceptible to salt and several antimicrobials.  In addition to increasing in response to cold, UFA levels also increased up to 11% in lag and exponential-phase cells grown at 20°C (Fig 4-7B), suggesting a role for UFAs in regular cell growth. In Saccharomyces cerevisiae, oleic acid and ergosterol supplementation mitigate oxidative stress (Landolfo et al., 2010). Similarly, bacteria transferred from a stationary-phase environment into a fresh oxygenated medium also experience oxidative stress (Rolfe et al., 2012) and may benefit from increased levels of membrane UFAs.  In Bacillus, phospholipid desaturases, encoded by des genes, are induced upon cold stress and controlled by a two-component regulator that senses changes in membrane fluidity (Aguilar et al., 2001; Diomande et al., 2016). These desaturases introduce double bonds at the Δ5 or Δ10 positions of acyl chains attached to existing phospholipids. Pseudomonas has a similar desaturase system that can introduce double bonds at the Δ9 position (Zhu, Choi et al. 2006). Currently, no lipid desaturases have been identified in Listeria. However, Vadyvaloo et al. (2002) showed that when L. monocytogenes is treated with a desaturase inhibitor, it exhibits decreased levels of membrane UFAs, demonstrating that such a system does exist. The synthesis of oleic (18:1 cis-Δ9) and palmitoletic (16:1 cis-Δ9) acids by L. monocytogenes suggests that it contains a desaturate system like that present in Pseudomonas. In the present study, we demonstrate the presence of  192  additional UFAs, that contain double bonds in several other locations (Supplementary Table 4-3) highlighting the possibility that more than one desaturase system exists in this bacterium.   Figure 4-8. Proportions of anteiso C15:0 and the combined sum of the unsaturated fatty acids 16:1 cis-Δ9 and 18:1 cis-Δ9 out of total FAs found in L. monocytogenes cells harvested across five growth phases at 4°C. Lipids were extracted from cell concentrates and resulting fatty acid methyl esters (FAME) were analyzed by gas chromatography. See Fig 4-1 for information about growth phases. 4.3.7 Comparison of antisense transcription at 20°C and 4°C A total of 92 - 261k paired-end (PE) reads were successfully mapped to the antisense strand of L. monocytogenes EGD ORFs, accounting for 0.61 - 1.63% of the total number of mapped reads.  Overall, 70% of genes had >10 mapped PE asRNA reads; however, only 56 genes had >1,000 asRNA mapped PE reads with a maximum of 40,000 PE reads observed for a LMON_R5 (vs. a maximum of 591,000 observed for a single mRNA transcript). These results are in line with  193  previous findings from other bacteria, from which asRNAs were detected for 20 - 75% of genes (Georg and Hess, 2011; Passalacqua et al., 2012; Sharma et al., 2010; Voigt et al., 2014).  Overall, more asRNA transcripts were upregulated than downregulated under cold stress (Fig 4-3). The largest number of upregulated asRNA transcripts was observed at G1 (Fig 4-3), which makes sense given that asRNAs act as gene regulators. At T1, cells are actively restructuring their transcriptional activity and cellular processes to prepare for growth in their new environment (Rolfe et al., 2012). Around 10% of the asRNAs currently described for L. monocytogenes are thought to be involved in regulating transcription regulators (Thomason and Storz, 2010). This idea is supported by the overall low levels of antisense expression observed in this study.  While many asRNA transcripts were upregulated at G1, total antisense transcription was generally highest in late stationary-phase cells at both temperatures (Table 4-8), highlighting the importance of antisense regulation at this growth phase. Compared with regulatory proteins, asRNA transcripts offer the advantage of being able to provide a rapid connection between quorum sensing and direct destabilization of target mRNAs; moreover, as they can act post-transcriptionally, they can more tightly control the repression of proteins under environmental stress conditions (Izar et al., 2011; Thomason and Storz, 2010). Finally, unnecessary regulatory RNAs can be cleared quickly and via a process that requires less energy consumption than the removal of regulatory proteins. This may in part explain the abundance of asRNAs during late stationary-phase, in which energy sources are limited.  Most genes with no or <10 asRNA transcripts encoded tRNAs (Supplementary Table 4-2). By contrast, rRNA had the highest levels of antisense transcription (Table 4-8). Similarly, Wehner et al. (2014) reported high levels of antisense transcription for rRNA in L. monocytogenes under intracellular growth conditions. Given the critical importance of ribosomes in cell functioning, it  194  makes sense that bacteria would take advantage of the tight means of regulation offered by antisense transcripts. Other highly expressed antisense transcripts targeted genes encoding the cell-shape determining protein, MreBH; flagellar biosynthesis protein, FliP; magnesium and cobalt transport protein, CorA; and two transcription regulators (LMON_0384, LMON_0738), among others (Table 4-8). Again, most of these genes exhibited maximum levels of asRNA transcription in stationary-phase cells with higher levels evident at 4°C than 20°C.  Some L. monocytogenes genes exhibited high levels of antisense transcription with no or very low levels of mRNA transcription. Not surprisingly, this occurred when an antisense transcript was an extension of a 3’ or 5’ untranslated region (UTR) of an adjacent gene (Fig 4-9ABF), suggesting that the asRNA probably prevented RNA polymerase from binding to the promoter region of the gene on the opposite strand. On the contrary, antisense transcripts that appeared to be individually transcribed, frequently had high levels of mRNA expression (Fig 4-9EG). These forms of asRNA have been shown to alter transcript stability by forming a sense/antisense RNA duplex leading to RNase-mediated degradation (Dühring et al., 2006; Lasa et al., 2012; Lee and Groisman, 2010) or by stabilizing mRNA transcripts by inducing cleavage of unstable polycistronic transcripts (Opdyke et al., 2011, 2004; Tramonti et al., 2008). Future research aimed at validating the mechanisms and functions of specific L. monocytogenes antisense transcripts will further enhance our understanding of gene regulation in this pathogen and possibly lead to the development of novel intervention strategies that can be used in the food industry.  195  Table 4-8. Top 20 most highly expressed antisense transcripts in L. monocytogenes at 20°C and 4°C. Overlapped EGD ORF Description of EGD ORF ORF strand ORF length (bp) Type of antisense transcript % of ORF covered Normalized paired-end read counts* C1 C2 C3 C4 C5 T1 T2 T3 T4 T5 LMON_0226 Hypothetical protein - 161 5’ UTRof LMON_0227 100 1693 1739 1200 730 618 738 1205 960 673 666 LMON_0347 Hypothetical protein + 113 3’ UTR of LMON_0348 62 75 49 54 118 593 120 89 109 370 3838 LMON_0384a Transcriptional regulator, DeoR family - 920 3' UTR of LMON_0383 100 421 291 415 952 6745 816 449 494 1467 21581 LMON_0652abc Magnesium and cobalt transport protein CorA + 950 5' UTR of  LMON_0651 38 33 17 42 874 2380 289 242 344 1411 12635 LMON_0681d Flagellar biosynthesis protein FliP + 767 IT 100 12 18 107 1715 5087 69 169 256 603 1644 LMON_0682d Flagellar biosynthesis protein FliQ + 272 IT 81 2 8 69 1642 2650 40 121 218 442 422 LMON_0738a Transcriptional regulator, XRE family + 509 3' UTR of LMON_0739 100 205 336 321 406 289 1044 574 612 590 17856 LMON_1143 Propanediol utilization transcriptional activator - 884 IT 100 239 120 818 1024 11556 222 106 176 241 364 LMON_1780a Cell-shape determining protein MreBH + 992 IT with LMON_T39 100 975 938 895 1403 2973 2575 1257 1311 1453 19114 LMON_2216 Hypothetical protein - 665 Unclear  100 5 3 3 55 460 16 6 3 39 3567 LMON_2351 Hypothetical protein - 443 3' UTR of pepC 100 115 108 183 307 1147 87 127 259 517 7613 LMON_2352 Pseudouridine 5'-phosphate glycosidase - 911 3' UTR of pepC 100 86 61 66 68 624 107 107 156 293 7949 LMON_2353 Pseudouridine kinase - 1118 3' UTR of pepC 100 61 48 43 37 442 102 81 98 179 6452  196  Overlapped EGD ORF Description of EGD ORF ORF strand ORF length (bp) Type of antisense transcript % of ORF covered Normalized paired-end read counts* C1 C2 C3 C4 C5 T1 T2 T3 T4 T5 LMON_2550 Hypothetical protein + 206 5' UTR of atpI 100 1932 3007 8743 17166 11120 631 2379 6384 15637 13462 LMON_2699ab ImpB/MucB/SamB family protein + 1256 3' UTR of LMON_2700 100 919 978 1240 1568 1422 873 841 968 1175 6461 LMON_R1 5S rRNA + 1520 IT 100 1515 2188 4954 8066 9881 5305 5839 4863 6117 16314 LMON_R2 23S rRNA + 2931 IT 100 3196 4648 10041 15825 19762 9955 11248 11037 14694 39471 LMON_R4 16S rRNA + 1520 IT 100 1561 2162 5015 8089 9955 5323 5919 4922 6091 16331 LMON_R5 23S rRNA + 2931 IT 100 3188 4578 10055 15765 19809 9819 11148 11018 14626 39964 LMON_T39 tRNA + 71 IT 100 73 73 71 170 255 193 105 113 146 1458 * The normalized read counts presented represent the average of the 2-3 biological replicates. ORFs with the highest abundance of antisense transcription were determined by dividing the number of PE reads per ORF by the length of each ORF. ORFs highlighted in yellow have been previously shown or are assumed in the present study to be long antisense transcripts covering multiple ORFs. The following superscripts denote the study in which a transcript was identified: a Wehner et al. (2014); b Wurtzel et al. (2012); c Mraheil et al. (2011); d Toledo-Arana et al. (2009). C1-C5 and T1-T5 represent the five growth phases at 20 and 4°C, respectively. Abbreviations: UTR = untranslated region; IT = individually transcribed 197   Figure 4-9. Coverage maps of select highly expressed antisense transcripts in L. monocytogenes cells grown at 20 or 4°C (refer to Table 4-6). The coverage maps shown are from stationary phase cultures grown at 4°C. Panels A-F shows individual paired-end, reads while panel G shows overall coverage trends for rRNA (5, 16, and 23S). Reads with the same start and end alignment positions are overlaid and depicted in green. Areas shaded in blue and orange denote sense and antisense transcripts of target genes, respectively.  198  4.4 Conclusions Here we present results from the first time-course study to investigate the transcriptional response and associated cold-induced membrane FA changes of L. monocytogenes, from early growth phases through to late stationary-phase. To the best of our knowledge this is also the first study to look at asRNA expression in L. monocytogenes during cold stress and across multiple growth phases. Our results revealed that the L. monocytogenes transcriptomic response to cold stress is most active during late stationary-phase survival. Similarly, we show that the L. monocytogenes cold stress regulon differs greatly across growth phases, highlighting the importance of carefully selecting appropriate time points when designing and conducting transcriptome studies.  Overall, more genes were suppressed than induced in L. monocytogenes under cold stress conditions. A core set of 22 genes was upregulated at all growth phases, including nine genes required for BCFA synthesis, the osmolyte transporter genes opuCBCD, and genes encoding the internalins A and D. This reflects the cell’s need to synthesize BCFAs to maintain membrane fluidity at low temperatures, and to support proper protein-folding through the uptake of structural-stabilizing osmolytes. Genes suppressed at 4°C were largely associated with cobalamin (B12) biosynthesis or the production/export of cell wall components. While is unclear how L. monocytogenes may benefit from downregulating cobalamin biosynthesis genes during cold stress, a reduced rate of peptidoglycan/cell envelope turnover at 4°C relative to 20°C may reflect the reduced cellular growth rate at this temperature, or allow carbohydrates to be conserved for alternative uses. Notable cold-induced membrane FA changes included a 15% increase in the proportion of BCFAs and a 15% transient increase in UFAs between the lag and exponential phases. Such  199  information may be useful for improving intervention strategies that target the food industry, as increased membrane UFA levels are known to increase the susceptibility of L. monocytogenes to salt and several antimicrobials. Overall, around 70% of L. monocytogenes genes exhibited antisense transcription but only a small portion of antisense transcripts exhibited expression levels comparable to that of mRNA. On average, antisense transcription was higher at 4°C than at 20°C, highlighting the importance of antisense regulation in the L. monocytogenes CSR. The largest number of upregulated antisense transcripts was observed during early lag phase; however, at both temperatures antisense transcription was generally highest in late stationary-phase cells. Stationary-phase cells likely benefit from the tight control of protein expression that is offered by antisense RNA, thereby reducing the overall energy requirement of the cell while faced with nutrient-limiting conditions. Collectively, our results reveal novel gene expression patterns and membrane FA alterations that occur in L. monocytogenes following cold stress, and highlight the abundance of antisense transcription in this microorganism at both 20°C and 4°C. We believe that this research will serve as a platform for future cold stress studies by facilitating the selection of candidate genes and antisense transcripts for further functional validation.   200  Chapter 5: Phenotypic and genetic characterization of Listeria monocytogenes enhanced cold tolerance variants isolated from prolonged cold storage cultures  5.1 Introduction  The human pathogen Listeria monocytogenes represents an ongoing concern in the food industry globally, where it is continuously detected in food products leading to costly recalls and loss of consumer trust. More concerning is that recently several North American outbreaks have implicated food products that have not traditionally been considered high risk for causing listeriosis, such as candy apples and ice cream (US CDC, 2017). The occurrence of these outbreaks demonstrates that we still have a great deal to learn about this foodborne pathogen and the factors facilitating its survival and/or growth in both food-processing environments and in foods.  While L. monocytogenes is most recognized for its ability to grow at refrigeration temperatures, it is also capable of tolerating a number of other food-related stresses such as growth in the presence of up to 12% salt or in media with a pH as low as 4.7 (Cole et al., 1990; Walker et al., 1990). However, bacterial strains are known to differ in their abilities to tolerate various stresses and these limits only represent the abilities of the specific strains that were evaluated in these studies. Differences between strain phenotypic behaviour commonly stem from the presence/absence of chromosomally or plasmid located genes, or from single nucleotide polymorphisms (SNPs) which can occur as random “’mistakes” during replication or as a result of a selective pressure such as antibiotics. While disadvantageous bacterial mutations are approximately 100,000× more common than beneficial mutations (~10-4 mutations per genome per replication vs. 10-9 in Escherichia coli) (Boe et al., 2000; Imhof and Schlotterer, 2001),  201  beneficial mutations remain a large concern for the food industry as arising strains may possess enhanced survival capabilities that render current intervention and safety measures ineffective. Given the importance of preventing L. monocytogenes from reaching unsafe levels in food, little is known regarding conditions that may select for the evolution of strains with enhanced abilities to tolerate food-related stresses. Prior to the rise and increased feasibility of whole genome sequencing, identifying advantageous stress-induced mutations in bacteria was extremely difficult but has now become a routine workflow. In a study by Karatzas et al. (2003), an L. monocytogenes Scott A high pr