Investigating the Role of Small Non-Coding Rnas in Aedes Aegypti-Ross River Virus Interaction

Investigating the role of small non-coding RNAs in Aedes aegypti-Ross River virus interaction

James Sinclair BBiomedSc (Hons)

0000-0003-4335-3183

A thesis submitted for the degree of Master of Philosophy at The University of Queensland in 2020 School of Biological Sciences Abstract

In mosquitoes, the activation of the RNA interference (RNAi) pathways, which involve a number of small non-coding RNAs (sncRNAs), is their predominant defence response to viral infection. The advent of comparative sncRNA expression analysis has enabled the discovery of genes involved in mosquito innate immunity responses to viral infection, however, the host-virus kinetics of the small RNAs involved in these pathways remains poorly characterised and there are limited studies characterising tissue-specific changes; particularly for changes in sncRNA expression induced by old-world alphaviruses. This project includes the first study of the sncRNA responses in mosquitoes to Ross River virus (RRV) infection.

RRV is the leading cause of arboviral disease in Australia and has been isolated from more than 40 mosquitoes but for most of these their role in transmission cycles or maintenance is unclear. Important vectors in Australia include Aedes vigilax, Aedes camptorhynchus, Aedes notoscriptus and Culex annulirostris. Aedes aegypti was putatively implicated in the large 1979-80 South Pacific Islands outbreak of RRV. In laboratory, Ae. aegypti competently transmits the virus but RRV has never been isolated from this mosquito in the wild. Marsupials such as Kangaroo or Wallaby are considered important reservoirs, but the field data is confounded by the involvement of many other vertebrate animals. There are no vaccines or therapeutic treatments for RRV and strategies to mitigate public health risk predominately involve mosquito control campaigns using insecticides. This thesis concerns an investigation of the defence attributes in the sncRNAs [short interfering RNA (siRNA), PIWI-interacting RNA (piRNA) and microRNA (miRNA)] produced by the RNA interference (RNAi) pathways that are involved in mosquito innate immunity against viral infection. The sncRNAs attributable to defence responses were identified by measuring their change in expression in Ae. aegypti midgut and fat body tissues resulting from an effect of the condition of infection with RRV using RNA sequencing. The method of comparative gene expression analysis used in this context enabled tissue-specific identification of differentially expressed miRNA candidates which were investigated for their putative involvement in Ae. aegypti defence responses and for the effect their modulation has on RRV replication. In contrast to using whole mosquitoes, the tissue-specific approach of this experimental design provides a clearer and more comprehensive characterisation of the unique set of molecular factors activated by the condition effect in different tissues.

Chapter 2 provides bioinformatic and laboratory-based evidence for the active sncRNA response in Ae. aegypti mosquitoes to RRV infection by characterising miRNA, siRNA, and piRNA responses. This work revealed that RRV induced an incremental RNAi response over time yielding

ii virus-derived short interfering RNAs (vsiRNA) and PIWI-interacting RNAs (vpiRNA). In addition, 14 host miRNAs were found to be differentially expressed due to RRV infection with the majority of those miRNAs recovered from the fat body. An analysis of the genes that were predicted as targets of those miRNAs showed that several were involved in innate immunity and defence response pathways. Conclusions drawn from those results provide an important and relevant contribution to our understanding of the molecular host-virus interactions within a medically significant vector infected with an old-world alphavirus.

Chapter 3 expands on the significantly differentially expressed miRNAs determined in Chapter 2 by investigating the pro-viral or anti-viral effect the modulation of five of these miRNAs may have on RRV replication in Ae. aegypti Aag2 cells. For miRNAs that were found to be significantly upregulated, the corresponding inhibitors of those miRNAs were transfected into Aag2 cells, and for those miRNAs that were significantly downregulated, mimics were used. Under these experimental conditions, cells were inoculated with RRV and the effect of miRNA increase or respective decrease in each treatment condition was investigated for that effect on RRV replication. While bioinformatics analyses indicated potential involvement of targets of these miRNAs in innate immunity and virus replication, experimental manipulation of the miRNAs tested did not have any effect on RRV replication.

Overall, work presented in this thesis illustrates the previously unexplored molecular dynamics between Ae. aegypti and RRV in the context of a tissue-specific host-virus model. The results of this project expand our understanding of insect innate immune responses to viral infection by providing greater insight into the repertoire and utility of Ae. aegypti sncRNA as a mechanism of defence.

iii Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, financial support and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my higher degree by research candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis and have sought permission from co-authors for any jointly authored works included in the thesis.

iv Publications included in this thesis

James Sinclair, and Sassan Asgari, Ross River virus provokes differentially expressed microRNA and RNA interference responses in Aedes aegypti mosquitoes, Viruses, 12, 695, 2020.

Submitted manuscripts included in this thesis

No manuscripts submitted for publication.

Other publications during candidature

No other publications.

Contributions by others to the thesis

Sassan Asgari conceived, supervised, administered, and acquired funding for this project and reviewed, edited, and advised on the methods and analyses presented in this thesis.

Statement of parts of the thesis submitted to qualify for the award of another degree

No works submitted towards another degree have been included in this thesis.

Research involving human or animal subjects

No animal or human subjects were involved in this research.

v Acknowledgments

I would like to acknowledge my principal supervisor Sassan for his tireless and unwavering energy and dependable support. You truly exemplified an open-door policy and whenever any solution seemed to elude me yours was always blazing quick and sometimes embarrassingly obvious. Thank you for this experience, and for all that you have done for me during my tenure, and for helping me to develop my science and its communication.

I am very grateful to my readers Karyn and Francesca and my co-supervisor Nigel for all your important and useful insights. Francesca Frentiu contributed with helpful advice on the methodological approach for this project. Nigel Beebe and Karyn Johnson contributed with important advice on the methods and analyses of this project and provided critical comments of the thesis and publication. Rhys Parry provided useful insight into the analysis and interpretation of research data in the publication.

A special thank you to the countless hours of useful conversations and guidance from my fellow ‘lab- lads’ Hugo, Cam, Sol, Lachie, and especially Rhys. I had come from a medical science background and you guys were fundamental in steering me through those early months and mentoring me through the many new procedures I needed to learn along the way.

On a personal note, my accomplishments were only made possible due to the continuous support and encouragement from my wife Marianne. Thank you for enabling me to pursue this dream and for your endurance!

vi Financial support

This research was supported by an Australian Government Research Training Program Scholarship.

Keywords

Aedes aegypti, Ross River virus, microRNA, short interfering RNA, PIWI-interacting RNA

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 060405, Gene Expression, 60% ANZSRC code: 060808, Invertebrate Biology, 30% ANZSRC code: 060506, Virology, 10%

Fields of Research (FoR) Classification

FoR code: 0601, Biochemistry and Cell Biology, 80% FoR code: 0699, Other Biological Sciences, 20%

vii Table of Contents

Abstract ...... ii Declaration by author ...... iv Table of Contents ...... viii List of Figures ...... x List of Tables...... xiii List of Abbreviations...... xiv

Chapter 1 Literature review ...... 1 1.1 Alphavirus range expansion and global impact ...... 1 1.2 Alphavirus genome organisation and replication ...... 2 1.3 Ross River virus importance and potential for emergence ...... 5 1.4 Vectors of Ross River virus ...... 7 1.5 Aedes aegypti prevalence and impact on the global community ...... 7 1.6 Mosquito barriers to viral infection ...... 8 1.7 Mosquito innate immunity ...... 9 1.8 RNA interference pathways ...... 12 1.9 Crosstalk between the innate immunity and the RNAi pathways...... 15 1.10 Thesis aims and hypothesis ...... 16

Chapter 2 RRV provokes Ae. aegypti miRNA and RNAi responses ...... 19 2.1 Introduction ...... 20 2.2 Materials and Methods ...... 23 2.2.1 Ethics statement ...... 23 2.2.2 Mosquito infections with Ross River virus ...... 23 2.2.3 Mosquito dissection and RNA extraction ...... 24 2.2.4 Library preparations and sequencing ...... 24 2.2.5 Small RNA analysis ...... 24 2.2.6 RNAi activity analysis ...... 26 2.2.7 miRNA target identification ...... 26 2.2.8 Gene Ontology ...... 26 2.2.9 RT-qPCR analysis ...... 27 2.3 Results and Discussion...... 27 2.3.1 Illumina sequencing of small RNAs ...... 27 2.3.2 Differential expression of Ae. aegypti miRNAs in response to RRV infection ...... 29 2.3.3 RT-qPCR validation of differentially expressed miRNAs...... 34 2.3.4 miRNA target analysis ...... 35 2.3.5 RRV is a target of the Ae. aegypti RNAi response ...... 38 2.3.6 Production of RRV-derived vpiRNAs ...... 39 2.4 Conclusions ...... 42

Chapter 3 Effect of miRNAs on RRV replication in Aag2 cells ...... 44 3.1 Introduction ...... 44 3.2 Materials and Methods ...... 45 3.2.1 Gene and microRNA differential expression analysis ...... 45 3.2.2 Reciprocally co-expressed microRNA and target gene analysis ...... 46 3.2.3 Gene ontology enrichment analysis of differentially expressed genes ...... 47 3.2.4 KEGG pathways analysis ...... 47 3.2.5 miRNA mimic/inhibitor transfection and Ross River virus inoculation ...... 47 3.2.6 RNA extraction and miRNA RT-qPCR ...... 48 3.2.7 RRV RT-qPCR...... 48 3.3 Results and discussion ...... 49 3.3.1 Correlation analysis of differentially co-expressed microRNAs and genes ...... 49 3.3.2 Gene Ontology enrichment analysis ...... 52 3.3.3 KEGG pathways enrichment analysis ...... 57 3.3.4 The effect of miRNA expression on RRV replication ...... 59 viii 3.4 Conclusion ...... 63

Chapter 4 General discussion ...... 65 4.1 Introduction ...... 65 4.2 Concluding remarks ...... 69

References ...... 70

ix List of Figures

Figure 1.1 The digitally rendered montage of cryoelectric scans of the RRV virion (upper left) displays the envelope proteins (blue-grey) and, by cut-out exposure, the host derived lipid bi-layer (tan) encasing the nucleocapsid core (red). The cross-section of the virion (centre and zoom) depicts the orientation of the envelope proteins [E1 (purple), E2 (aqua- grey), and E3 (orange)] and how they span the lipid-bilayer and integrate internally with the capsid proteins (box: blue) that line the nucleocapsid core. The RNA genome is located within the nucleocapsid core. Composition of different images adapted from various sources (Strauss and Strauss, 1994; Weaver and Lecuit, 2015; Zhang et al., 2011)...... 3

Figure 1.2 The alphavirus genome (central) has two open reading frames (ORF) that encode the non-structural and structural proteins. Opal termination (black triangle) provisionally translates an early P123 and a late P1234 polyprotein following read-through. The non-structural polyproteins are processed by nsP2 protease activity yielding negative-sense (P123+nsP4) and positive-sense (nsP1+nsP2+nsP3+nsP4) replicases by which the negative-sense viral complementary RNA (vcRNA) and complete positive-sense genomic RNA (gRNA) are transcribed. mRNA transcribed from a 26S subgenomic promotor is translated into the structural polyprotein which is further processed to produce the individual structural proteins. Figure adapted from (Strauss and Strauss, 1994)...... 5

Figure 1.3 Schematic of the innate immune pathways Toll, IMD, JAK-STAT, and RNAi in mosquitoes (Sim, et al. 2014)...... 10

Figure 1.4 The major small non-coding RNAs (sncRNAs) of the mosquito RNA interference (RNAi) pathway involves, (A) short interfering RNA (siRNA), (B) microRNA (miRNA) and, (C) P-element induced wimpy testis RNA (piRNA). Figure adapted from (Hussain et al., 2016a)...... 12

Figure 2.1 Measurement of Ross River virus (RRV) infection in Aedes aegypti mosquitoes at 2, 6, and 12 days post- infection (dpi) by RT-qPCR. (A) Rate of infection: percentage of mosquitoes with a detectable degree of RRV infection (N = 10), and (B) their relative viral load (N = 6). Error bars indicate the standard error mean of the mean normalized expression values. The Ae. aegypti ribosomal protein subunit 17, RPS17, gene was used as the internal calibrator...... 28

Figure 2.2 Length distribution of quality-filtered and adapter trimmed reads from Aedes aegypti fat body and midgut tissues at 2, 6 and 12 dpi. Error bars indicate the standard error mean of the biological replicates (N = 3)...... 29

Figure 2.3 RRV causes differential expression of Aedes aegypti miRNAs. Log2 fold changes of miRNA expression following infection with RRV at 2 dpi in fat body (green) and midgut (yellow) and in fat body at 12 dpi (purple). Error bars represent the standard error mean of biological replicates (N = 3)...... 30

Figure 2.4 RRV caused differential expression of Aedes aegypti miRNAs mostly early in infection. Volcano plots show miRNA log2 fold changes due to RRV infection in fat body at (A) 2 dpi, (B) 6 dpi, and (C) 12 dpi, and in midgut at (D) 2 dpi, (E) 6 dpi, and (F) 12 dpi. Significantly differentially expressed miRNAs (orange) were found in fat body samples at 2 dpi and 12 dpi and in midgut samples at 2 dpi...... 30

Figure 2.5 Aedes aegypti miRNA differential expression due to RRV infection between tissues and across time. The expression profile of (A) miR-9b-5p and (B) miR-71-3p, in control (blue) and infected (red) fat body and midgut samples at 2, 6, and 12 dpi. Significant changes are labelled with the fold change (FC) and adjusted p-value (padj). The boxplot and whiskers show the mean of the normalized reads counts of the biological replicates (N = 3)...... 33

Figure 2.6 RRV induces differential expression of Aedes aegypti miRNAs. Contrasting RNA-Seq and RT-qPCR measurement of RRV-induced differential expression of 10 miRNAs in Ae. aegypti fat body and midgut RNA samples at 2 dpi. Error bars represent the standard error mean of the biological replicates (N = 3)...... 34

Figure 2.7 Venn diagram showing the number of predicted targets for each miRNA and the intersect shows the number of predicted targets in common...... 35

Figure 2.8 Molecular function. Tree map of GO descriptions ascribed to genes from this study showing the interrelationship between the parent (blue) and children (shades of orange) ontologies involved in the molecular functions of binding and catalytic activities...... 36

x Figure 2.9 Molecular function. The number and percentage of genes from this study ascribed to various molecular functions...... 37

Figure 2.10 The proliferating 21 nt short interfering RNA (siRNA) count over time is indicative of a progressively increasing viral replication and RNAi response in Aedes aegypti mosquitoes. These time-series plots show the (A) length distribution of reads mapping to the RRV (T48) genome at 2, 6, and 12 dpi, and (B) the 21 nt siRNA coverage of sense (blue lines) and anti-sense (red lines) RRV (T48) genome strands. Error bars represent the SEM of biological replicates (N = 3)...... 39

Figure 2.11 RRV-derived PIWI-interacting RNAs (piRNAs) generated by the fat body and midgut tissues of Aedes aegypti mosquitoes. Distribution of 27-29 nt small RNAs that mapped across the sense (blue) and anti-sense (red) strands of the RRV genome at 2, 6 and 12 dpi. Distinct 3’ hotspots indicate virus-derived (vpiRNA) production against subgenomic RNA...... 40

Figure 2.12 RRV vpiRNA signatures in Aedes aegypti. (A) There was an A10 nucleotide bias, indicative of ping-pong amplification, present in the 28 nt sense vpiRNA-like strands produced in Ae. aegypti fat body tissue at 6 dpi with RRV, although there are other biases at positions 1, and 26-28. (B) The U1 bias, a signature of piRNA ping-pong amplification, was evident in 28 nt piRNAs produced in Ae. aegypti fat body at 6 dpi...... 41

Figure 2.13 Sequences in the range 27-29 nt from Aedes aegypti fat body at 6 days post infection with RRV exhibit a high 10 nt overlap probability. (A) A tally of sequence pairs by the number of their overlapping nucleotides, and (B) the probability that sequence pairs overlap by the number of intersecting nucleotides. The z-scores (axis on the right) show the mean +/- the number of standard deviations away from the mean. The number of sequences that overlap by 10 nucleotides and the probability that sequence pairs overlap by 10 nt are higher than two standard deviations from the mean. These peaks at 10 nt are a signature of piRNA production by ping-pong amplification...... 42

Figure 3.1 Comparing the respective sums of significantly differentially depleted (blue) or enriched (yellow) microRNAs (top, miRNAs) and Genes (bottom) that were found in Ross River virus infected Aedes aegypti midgut or fat body samples at 2, 6, and 12 days post infection (dpi). An analysis comparing the respective profiles of expressional change showed very similar trends in the ratio of the number of respective genes that was differentially expressed at any given day with the majority of changes commonly occurring at 2 dpi. Additionally, both groups had relatively few instances of changes in gene expression in fat body tissues at 12 dpi and neither group had any significantly differentially expressed miRNAs at 6 dpi or genes in midgut tissues at 6 dpi. There were contrasting differences in the overall directions of change between groups in the fat body at 2 dpi, with the ratio of differences especially pronounced in the fat body at 2 dpi...... 50

Figure 3.2 Venn diagram showing the major interactions of the significantly differentially expressed (SDE) genes found in Aedes aegypti the fat body and midgut samples following infection with Ross River virus. Excluding the midgut samples at 6 dpi, SDE genes were found in both the fat body and midgut samples at 2, 6, and 12 dpi. Shown in the intersecting pink and purple circles are the counts of the predicted miRNA target genes (labelled ‘targets’) of the SDE miRNAs found in the fat body and midgut samples at 2 dpi (determined in Chapter 2). The target gene predicted for the fat body samples at 12 dpi is not shown. Shown in bold are the number of SDE genes common among samples...... 51

Figure 3.3 The grids of the volcano plots show the significant differentially expressed (SDE) enrichment (red) or depletion (blue) of the microRNAs (left) and genes (right) in Ross River virus infected Aedes aegypti fat body (top) and midgut (bottom) samples at 2 dpi. A minimum of 3/5 software programs predicted that the labelled miRNAs of each respective tissue would target at least one of the corresponding genes. The thresholds for significant differential expression was an adjusted p-value (Benjamini and Hochberg) < 0.1 and fold changes greater than 1.5. Results below these thresholds (grey) were considered as not significant (NS) and genes in this category that were predicted as a miRNA target were ignored. Of particular interest is the enriched miRNAs (red) that were predicted to target depleted genes (black). Legend terms: SDE (significant differential expression), DOWN (significant down-regulation), UP (significant up-regulation), UD (miRNA up-regulated – target gene down-regulated), DU (miRNA down-regulated – target gene upregulated), UU (both miRNA and target gene up-regulated), DD (both miRNA and target gene down-regulated), NS (not significant)...... 52

Figure 3.6 Biological process Gene Ontology terms of Aedes aegypti fat body genes at 2 dpi with Ross River virus. The number of annotated genes were found to be significantly differentially expressed with p-values < 0.01. The topGO analysis in which results were determined to be significantly enriched are shown in blue with Wilcoxen test p-values shown in white. Results that were not determined to be enriched are here labelled as not significant (ns)...... 56

Figure 3.7 KEGG pathways enrichment of differentially expressed genes in Ross River virus infected Aedes aegypti fat body and midgut samples at 2 dpi. The numbers above each column are the counts of annotated genes. The dashed red lines, demarcating the p-values 0.1 (lower) and 0.05 (upper), represent the thresholds of significance as determined by the Wilcoxon test...... 58

Figure 3.8 KEGG pathways of differentially expressed miRNA target genes in Ross River virus infected Aedes aegypti fat body and midgut samples at 2 dpi. The numbers above each column are the counts of annotated genes of each respective pathway. The dashed red line indicates the threshold of significant enrichment (p-value 0.1) as determined by the Wilcoxon test. In this case, only the mitogen activated protein kinase (MAPK) KEGG pathway was found to be enriched. This group of 8 pathways were selected from among a total of 12 pathways for which fat body or midgut genes at 2 dpi were annotated. The other pathways which were omitted from this display for not being significant or directly related to a defence response, were all from the fat body samples at 2 dpi and included Fatty acid elongation, Biosynthesis of unsaturated fatty acids, Glycolysis / gluconeogenesis, and Starch and sucrose metabolism...... 59

Figure 3.9 Aedes aegypti cells (Aag2) infected (red) with Ross River virus (RRV) following transfection (blue) with the mimics (MMC) or inhibitors (INH) of a select number of microRNAs (miRNAs) that were determined in Chapter 2 to be respectively depleted or enriched in Ae. aegypti fat body tissues following infection with RRV. Relative quantification of miRNA or RRV levels was measured as the mean normalised expression (MNE) by RT-qPCR using the respective Ae. aegypti U6 or RPS17 gene as the calibrator. A one-way analysis of variation (ANOVA) was used with a multi-way comparison for the statistical test between the mean counts of respective columns for each set (Transfected or Infected). For treatment controls, Cellfectin only (CF) and the respective negative controls (NC) for the mimic or inhibitor treatments were used. The respective levels of significance are represented by p-values less than 0.001 (***) or 0.0001 (****). Points per column represent the test replicates of cultured repeats (n = 3) with error bars showing the standard error mean (SEM). Only significant differences between respective columns are indicated and all other comparisons of difference within each set ought to be inferred as not significant...... 61

xii List of Tables

Table 2.1 Differentially expressed Aedes aegypti miRNAs upon RRV infection...... 31

xiii List of Abbreviations

Aa Amino acid ADF Australian Defence Force Ae Aedes AeMMP1 Matrix metalloproteinase 1 AeMMP2 Matrix metalloproteinase 2 AGO Argonaute apeglm Approximate posterior estimation for generalized linear model ARC Australian Research Council Aub Aubergine BED Browser extensible data BLAST Basic local alignment search tool bp Base pair C or con Control CHIKV Chikungunya virus CHS Chalcone synthase cnd Condition CPE Cytopathic effect db Database Dcr Dicer DE Differentially expressed DENV 1 Dengue virus serotype 1 DENV 2 Dengue virus serotype 2 DENV 3 Dengue virus serotype 3 DENV 4 Dengue virus serotype 4 DNA Deoxyribonucleic acid DNase Deoxyribonuclease Dome Domeless dpf Day/s post feeding dpi Day/s post inoculation dsRNA Double-stranded ribonucleic acid EDTA Ethylenediaminetetraacetic acid et al And others FB Fat body FC Fold change FDR False discovery rate FHV Flock House virus G3BP1 Ras GTPase-activating protein-binding protein 1 G3BP2 Ras GTPase-activating protein-binding protein 2 GEO Gene expression omnibus GNBP Gram-negative bacteria-binding protein GO Gene ontology h Hour HSPG Heparin sulphate proteoglycan Imd Immune deficiency JAK Janus Kinase JNK c-Jun amino-terminal kinase JUN Jun proto-oncogene, AP-1 transcription factor subunit kcal Kilo calorie

xiv KEGG Kyoto encyclopedia of genes and genomes lncRNA Long non-coding RNA MMP Matrix metalloproteinase mb Megabase MEB Midgut escape barrier mfe Minimum folding energy MG Midgut MIB Midgut infection barrier min Minute miRNA Micro ribonucleic acid mL Millilitre mM Millimolar MM Mitsuhashi and Maramorosch medium MNE Mean normalized expression MOI Multiplicity of infection mRNA Messenger ribonucleic acid NaCl Sodium Chloride NaOAc Sodium acetate NCBI National center for biotechnology information ncRNA Non-coding RNA NF-kB Nuclear factor kappa B subunit ng Nanogram NGS Next-generation sequencing nm Nanometre nr Non-redundant dNRAMP Drosophila natural resistance-associated macrophage protein 2 nsP1 Non-structural protein 1 nsP2 Non-structural protein 2 nsP3 Non-structural protein 3 nsP4 Non-structural protein 4 nt Nucleotide ORF Open reading frame padj Adjusted p-value PAMP Pathogen associated molecular pattern PBS Phosphate buffered saline PCR Polymerase chain reaction piRNA Piwi-interacting ribonucleic acid piRISC Piwi-interacting ribonucleic acid induced silencing complex Poly(A) Polyadenylation pre-miRNA Precursor micro ribonucleic acid pri-miRNA Primary micro ribonucleic acid PRR Pathogen recognition receptor p-value Calculated probability Rel Relish REST RE1 silencing transcription factor RH Relative humidity RISC RNA-induced silencing complex RNA Ribonucleic acid RNA-Seq Ribonucleic acid sequencing RNAi Ribonucleic acid interference RNAPII Ribonucleic acid polymerase II rRNA Ribosomal ribonucleic acid

xv RRV Ross River virus RT Room temperature SGEB Salivary gland escape barrier SDE Significantly differentially expressed SDM Schneider’s Drosophila Medium sec Second SFV Semliki Forest virus SGIB Salivary gland infection barrier siRNA Short interfering ribonucleic acid sncRNA Small non-coding ribonucleic acid snoRNA Small nucleolar ribonucleic acid SRA Sequence read archive STAT Signal Transducer and Activator TE Transposable element tRNA Transfer RNA µg Microgram µL Microlitre tsu Tissue UQ University of Queensland upd Unpaired UTR Untranslated region V Virus infected vpiRNA Virus-derived Piwi-interacting ribonucleic acid YFV Yellow Fever virus ZIKV Zika virus Zuc Zucchini

xvi Chapter 1 Literature review

Literature review

1.1 Alphavirus range expansion and global impact

Alphaviruses (Togaviridae) are globally distributed and predominately mosquito-borne pathogens that cause encephalitic [e.g. Eastern (EEEV), Western (WEEV), and Venezuelan Equine Encephalitis (VEEV) viruses] or arthritogenic [e.g. Ross River virus (RRV), Chikungunya virus (CHIKV), Sindbis virus (SINV)] diseases in both humans and animals (Contigiani and Diaz, 2017). The Alphavirus genus has 31 members divided into old-world (e.g. RRV, CHIKV, SINV) or new- world (e.g. EEV, WEEV, VEEV) categories based on phylogenetic origin (Lefkowitz et al., 2018; Strauss and Strauss, 1994). Alphaviruses are maintained in sylvatic transmission cycles (animal- mosquito-animal) between amplifying vertebrates that develop a high and long-term titre of viraemia from the infected saliva of a biting mosquito (Agarwal et al., 2017). Occasionally, short-term amplifying hosts such as peridomestic animals facilitate outbreaks in semi-urban transmission cycles and, during epidemics, transmission cycles that are entirely urban (human-mosquito-human) can occur in densely populated regions (Koolhof and Carver, 2017). Due to drivers such as urbanisation, increased human travel, or the adaptive mutations of a virus to a new mosquito vector, the geographical range of alphaviruses, and, subsequently, the number of cases of disease they cause, are increasing (Callender, 2018; Hennessey et al., 2016; Mayer et al., 2017; Wahid et al., 2017).

The alphavirus RNA genome rapidly evolves and is capable of adapting to a broad range of mosquito hosts that may have very distinct ecological niches. For example, CHIKV outbreaks are usually facilitated by Ae. aegypti in tropical climates, whereas, the emergence of a strain of CHIKV (A226V), with increased fitness for transmission by Aedes albopictus, was responsible for outbreaks in the temperate regions of La Réunion in 2005 (Vazeille et al., 2007) and Italy in 2007 (Rezza, 2014). A large epidemic of RRV that spread throughout the Pacific Islands during 1979-80 was likely facilitated by a viraemic Australian tourist to Fiji (Aaskov et al., 1981; Russell, 2002b) and since then cases of autochthonous transmission together with seroepidemiological evidence suggest that circulation of RRV in Fiji has occurred endemically (Aubry et al., 2017; Lau et al., 2017). The strain of RRV (F9073) that was involved in and emerged in Australia sometime during the decade prior to the 1979-80 outbreak likely infiltrated Fiji by a viraemic Australian tourist. F9073 featured a duplication mutation in the hypervariable domain of the non-structural protein 3 (nsP3) gene, that 1 was rapidly fixed and all strains of RRV without that mutation have since become extinct (Aaskov et al., 2011). This suggests that the duplication mutation caused a competitively advantageous gain in fitness that led to the displacement of single-copy lineages, although, because there are no preserved samples in existence, a direct comparison of difference in fitness between strains cannot be made. Furthermore, a hypothesis for the duplication mutation being a factor in potentiating the 1979-80 outbreak, remains conjectural. The increasing prevalence of unforeseen outbreaks of disease caused by the globalisation of alphaviruses necessitates a greater understanding of the molecular interaction between these and their principal or potential vectors to elucidate the mechanisms of adaptation with the overarching goal to predict, prepare for or prevent such events.

1.2 Alphavirus genome organisation and replication

An alphavirus (45-75 nm in diameter) is an encapsidated positive-sense single-stranded RNA genome ( 12000 nt) encased in a host-derived lipid bilayer studded with envelope glycoproteins (Strauss and Strauss, 1994) (Figure 1.1). There are 240 copies of each glycoprotein (envelope proteins: E1 and E2) anchored in the lipid bilayer (Paredes et al., 1993). The E1 proteins comprise fusion peptides that form an icosahedral lattice with interface residues in the virion (Paredes et al., 1992). The architecture of E2 in coupling with E1 resembles a protruding spike that functions in receptor attachment (e.g. SINV: NRAMP2 (Rose et al., 2011)), a process required for virion entry into the cell (Garoff and Simons, 1974). Following clathrin-mediated endocytosis a pH-induced change in E1 protein conformation promotes fusion between viral and endosome lipid envelopes resulting in the liberation of the viral genome into the cytoplasm (Li et al., 2010).

The capped 5’ and polyadenylated 3’ genome is bicistronic encoding four non-structural (nsP1, nsP2, nsP3, and nsP4) and six structural (capsid, E1-3, 6K, and TF) proteins, respectively (Strauss and Strauss, 1994). Resembling endogenous mRNA, the viral genome is quickly translated by target cell ribosomes and other cell proteins (e.g. cap binding complex, G3BP1, G3BP2) (Montgomery et al., 2006). An opal stop codon (UGA) at the C-terminus of the nsP3 gene determines the polyprotein translation of P123 or, following read-through, P1234 (figure 1.2). P1234 represents ~10% of translations and during the early phase of replication P123 polyproteins accumulate in abundance. In vertebrate cells, the nsP2 proteins of many old-world alphaviruses [e.g. SINV and Semliki Forest virus (SFV)] detach from P123 by autocleavage, migrate into the nucleus, bind and rapidly degrade the host’s RNA polymerase II subunit rpb1 (Garmashova et al., 2007). Binding rpb1 causes the inhibition of global transcription (transcriptional shutoff) and blockage of alpha/beta interferon (IFN- 훼/훽) signalling (Akhrymuk et al., 2018a; Fros and Pijlman, 2016) resulting in host-cell cytopathic effects (CPE) (Sadler and Williams, 2008) and cell death by apoptosis within 24-48 h. Alphaviruses 3 replicate quickly and by 2 hours post-infection (hpi) levels of vertebrate cell rpb1 are less than 60% and by 8 hpi rpb1 levels are completely reduced. Conversely, the capsid proteins of many new-world alphaviruses impede host-cell nuclear import by blocking nuclear pores which also results in transcriptional shutoff in vertebrate cells (Atasheva et al., 2010). Recently, it was shown that SINV nsP3-specific mono-ADP-ribosylhydrolase activity induces translational shutoff and is a contributing factor of CPE in vertebrate cells by an as yet poorly understood mechanism (Akhrymuk et al., 2018b). Thus, diverse species of alphaviruses have evolved distinct mechanisms to disarm the molecular defense responses of vertebrates to the same effect. Alphaviruses do not cause transcriptional or translational shutoff in mosquitoes and so a persistent infection ensues with little to no CPE. Following the initial disarmament of the vertebrate host-cell defence system, read-through of the opal stop codon enables the late translation of P1234. The protease activity of nsP2 directs the proteolytic cleavage of the non-structural protein components of P1234, which reassemble to form replication complexes (replicase) comprising distinct conformations that direct the subgenomic transcription of the minus and plus strand 49S mRNA (Figure 1.2). The 26S subgenomic mRNA forms the template from which the structural proteins are translated (Strauss and Strauss, 1994).

Replication occurs in the cytoplasm within cell membrane invaginations called spherules (Grimley et al., 1972). The two-stage assembly mechanism model suggests that nucleocapsid packaging proceeds when capsid proteins align along the newly replicated genome and then systematically converge with one another to compress and internalise the genome within the nucleocapsid core (Mendes and Kuhn, 2018). The envelope proteins E1 and E2 are processed by the endoplasmic reticulum (ER) and delivered to and integrate with the nucleocapsid core that becomes sheathed in a cell derived lipid envelope during exocytosis (Brown et al., 2018).

1.3 Ross River virus importance and potential for emergence

RRV is endemic in Australia, Papua New Guinea, and Fiji and causes epidemic polyarthritis (EPA; otherwise known as Ross River fever or disease), the most prolific viral disease in Australia (NNDSS, 2018). EPA is characterised by a debilitating polyarthritis in the small joints of the

5 extremities and is often accompanied by a maculopapular rash on the limbs and trunk (Harley et al., 2001; Suhrbier and La Linn, 2004). RRV has multiple and complex transmission cycles which may be entirely sylvatic, urban, or a mixture of sylvatic and urban (Harley et al., 2001). Examples of the rapid spread of RRV infection propagated by urban (human-mosquito-human) transmission cycles include the aforementioned Pacific Islands outbreak of 1979-1980 (Aaskov et al., 1981), the outbreaks in Perth, and Western Australia in 1988-1989 and 1991-1992 (Lindsay et al., 1992), the outbreaks in Brisbane, Queensland in 1992 and 1994 (Ritchie et al., 1997), and the outbreaks among Australian Defence Force (ADF) personnel in Shoalwater Bay, north-eastern Australia, in 2016 and 2017 (Liu et al., 2019).

Many vertebrates are alleged to be involved in maintaining RRV (Claflin and Webb, 2015) but there remains a paucity of field data to substantiate any clear role these animals may have in RRV maintenance or amplification (Stephenson et al., 2018). Macropodid marsupials such as the grey kangaroo (Macropus giganteus) or agile wallaby (M. agilis) are widely considered the principal reservoir hosts of RRV, due to their high and prolonged titre of viraemia (Stephenson et al., 2018), however, their definitive role, confounded by the putative involvement of many other animals, remains unverified. Furthermore, Togami et al (2020) provided serological evidences of the endemic transmission of RRV in Fiji where there are no marsupials (Togami et al., 2020). Horses (Equus caballus) infected with RRV develop clinical signs of musculoskeletal disease of which there is a high and prolonged titre of viraemia (Azuolas et al., 2003), however, their role in transmission cycles are poorly defined (Gummow et al., 2018). Australian brushtail possums (Trichosurus vulpecula) also develop a high titre of viraemia (Boyd et al., 2001) and are circumspectly associated with being hosts during outbreaks due to their close proximity to humans living in metropolitan areas (Old and Deane, 2005), however, definitive data remains wanting. Other animals capable of transmission include chickens, dogs, pigs, birds, rabbits, and rats (Claflin and Webb, 2015). In Australia, grey headed flying foxes were found to be potentially involved in transmission cycles (Ryan et al., 1997) and in China RRV was detected in bats (Xia et al., 2018). In Australia, the putative native and domestic animal reservoirs have not been well defined and so the relative role(s) of the virus, the vector(s) and the host(s) remain unclear (Claflin and Webb, 2015). The unusually large number of hosts and vectors of RRV calls for a recognition of the significant potential for this virus to expand into regions with or without marsupials. The range expansion of RRV into the Pacific Islands and the possible endemic circulation of RRV even in China (Chunsheng et al., 1997; Xia et al., 2018; Zhao et al., 2000) are testaments of the significant potential for RRV range expansion.

6 1.4 Vectors of Ross River virus

RRV was first isolated from Aedes vigilax near Townsville, Queensland in 1959 (Doherty et al., 1963) and since then RRV has been isolated from 42 species of mosquito, whereas, laboratory competence studies have only indicted 10 as being capable of transmission (Russell, 2002b). The most epidemiologically important species in Australia include the saltmarsh mosquitoes Ae. vigilax in the coastal regions of the North (Whelan et al., 1993) and Ae. camptorhynchus in the South (Russell, 2002b). Culex annulirostris is an important inland freshwater species (Russell, 1994) and Ae. notoscriptus in urbanised regions (Ritchie et al., 1997). Interestingly, Ae. notoscriptus is associated with transmission cycles involving the Australian brushtail possum (Trichosurus vulpecula Kerr) (Boyd et al., 2001; Kay et al., 2007), both of which are prevalent in New Zealand (Derraik, 2004; Holder et al., 1999). Although there are no records of RRV circulation in New Zealand, the presence of Ae. notoscriptus and brushtail possums indicates a significant potential for local transmission if RRV were to be introduced (Kelly‐Hope et al., 2002). There is no definitive record of the species of mosquitoes involved in the Pacific Islands epidemic of 1979-1980, whereas, those implicated based on prevalence and capacity to readily transmit RRV included Ae. vigilax, Cx. annulirostris, Ae. polynesiensis, Ae. oceanicus, Cx. quinquefasciatus, and Ae. aegypti (Gubler, 1981; Nasci and Mitchell, 1994).

1.5 Aedes aegypti prevalence and impact on the global community

Ae. aegypti is unrivalled in medical importance being culpable for transmission of the most impactful viruses including, but not limited to, CHIKV, DENV, Yellow Fever virus (YFV), and ZIKV, wherein, collectively, tens of millions of cases of disease and tens of thousands of deaths occur globally each year (Bhatt et al., 2013; Musso and Gubler, 2015). Epidemiologic and phylogenetic evidence suggest that Ae. aegypti competently transmits CHIKV, ZIKV, DENV and YFV because these had a common African ancestry and subsequently coevolved (Powell, 2018). During the 15th and 16th century, slave traders deporting Africa by ship had inadvertently trafficked the anthropophilic strain of Ae. aegypti into the New World (Tabachnick, 1991) and due, to increased human travel, Ae. aegypti is now globally dispersed (Kraemer et al., 2015). The medical importance of Ae. aegypti is underscored by its capacity to adapt to a range of ecological niches, both spatial and temporal, to transmit multiple, disparately related, and highly pathogenic viruses, and for its natural drive to preferentially seek out human blood.

7 1.6 Mosquito barriers to viral infection

Viraemic blood imbibed by a mosquito is delivered into the midgut lumen which stimulates the midgut epithelium to release peritrophic matrix, a proteoglycan matrix that sheaths the blood meal in preparation for digestion. The barricading effect of peritrophic matrix represents the midgut infection barrier (MIB), whereby, virions suspended in the blood meal must aggregate with and infect the midgut epithelial cells prior to the release of the peritrophic matrix in order to bypass this first barrier to infection. In Ae. aegypti, peritrophic matrix is formed between 4-8 hours post blood feeding (Perrone and Spielman, 1988). The viral progeny that emerges from infected epithelial cells must now traverse the basal lamina which lines the posterior aspect of the epithelial tissue and represents the midgut escape barrier (MEB). Until recently, the mechanism whereby a virion could traverse the basal lamina has been unknown. The basal lamina is reinforced by collagen IV and in a CHIKV/Ae. aegypti model it was recently shown that the stretching of abdominal tissue resulting from the ingestion of a blood meal causes collagen IV degradation, which stimulates the production of matrix metalloproteinases (MMPs) AeMMP1 and AeMMP2 (Dong et al., 2017a), that remodel and repair collagen IV (Page-McCaw, 2008; Stevens and Page-McCaw, 2012). It is during this remodelling phase that virions are inadvertently provided a window of opportunity to traverse the basal lamina and disseminate. Virions that escape the midgut enter the haemocoel and disseminate to various tissues including fat body, haemocytes, nerve and muscle tissue, whereas, to be successfully transmitted they must multiply and reach the salivary glands (Franz et al., 2015). Haemocytes are an important vehicle in which to amplify and escort virions to the salivary glands (Parikh et al., 2009; Romoser et al., 2004). This ‘trojan horse’ concept is reflected in human infections. Chemicals in mosquito saliva stimulate the release of histamines (Ohtsuka et al., 2001) that cause oedema (Dietzel et al., 1969) which corrals viruses at the site of inoculation (Pingen et al., 2016). Oedema signals the cascading recruitment of inflammatory immune cells (Scapini et al., 2000) that counterintuitively become the vehicles of viral dissemination (Marovich et al., 2004).

Both mosquito salivary glands have three lobes and the central ducts within each lobe are lined with epithelial cells externally wrapped in basal lamina (Franz et al., 2015). Similar to the MEB, the basal lamina here represents the salivary gland infection barrier (SGIB), which has been hypothesised to occlude virion attachment to central duct epithelial cell receptors (Romoser et al., 2005). The final impediment to successful transmission is the salivary gland escape barrier (SGEB). Although many studies provide evidence of SGEB, sighting a bottleneck at the salivary glands (Forrester et al., 2014) whereby relatively few virus particles breach into saliva, the underlying mechanisms of these phenomena remain hypothetical. One mechanism proposed to explain breaching requires virions to

8 selectively cause apoptosis only in the gland acinar cells (salivary gland epithelial cells) of the lumen resulting in their break-through into saliva (Kelly et al., 2012). Interestingly, several studies have linked the viral induced cytopathic effects (CPE) in mosquito salivary gland cells with modified glandular secretions and increased appetite (Bowers et al., 2003; Kelly et al., 2012; Platt et al., 1997). Ciano et al (2014) proposed that heparan sulphate proteoglycan (HSPG) found in the filamentous membrane extensions in the internal ducts of salivary gland lobes were integral virus attachment factors and that CPE therein is associated with viral passage into saliva (Ciano et al., 2014).

1.7 Mosquito innate immunity

Drosophila melanogaster has been the vanguard model of cellular eukaryotic biology for more than a century and has strongly influenced research into mosquito immunity. Sequencing the genomes of D. melanogaster (Adams et al., 2000), Anopheles gambiae (Holt et al., 2002), Ae. aegypti (Nene et al., 2007), and Culex quinquefasciatus (Arensburger et al., 2010), among many other dipterans, has provided a platform for comparative genomic analysis (Fragkoudis et al., 2009). Genomic studies analysing differentially regulated genes has enabled significant progress in elucidating mosquito innate immune responses to arbovirus infection. Sim et al (2005) pioneered this approach reporting the modulation of An. gambiae gene expression in response to O’nyong-nyong virus (ONNV: Alphavirus) infection and later (2007) demonstrated that this modulation resulted in the production of heat shock protein cognate 70B (HSC70B), which impeded ONNV replication (Sim et al., 2005; Sim et al., 2007). Features of the innate immunity, including the Toll, Immune deficiency (Imd), c- Jun amino-terminal kinase (JNK), and Janus Kinase/Signal Transducer and Activator of Transcription (JAK/STAT) pathways (Figure 1.4) are highly conserved between D. melanogaster and Ae. aegypti orthologs (Waterhouse et al., 2007). In response to viral infection, Ae. aegypti has been reported to initiate Toll (Xi et al., 2008), JNK (Sanders et al., 2005), Imd (Luplertlop et al., 2011), and JAK/STAT pathways (McFarlane et al., 2014).

Figure 1.3 Schematic of the innate immune pathways Toll, IMD, JAK-STAT, and RNAi in mosquitoes (Sim, et al. 2014).

The canonical JAK/STAT pathway is activated when the extracellular glycoprotein, Unpaired (Upd), attaches to and dimerizes the transmembrane receptor Domeless (Dome), which results in the phosphorylation of associated JAKs that subsequently attract STAT proteins (Harrison et al., 1998). The Dome/JAK complex phosphorylates the STAT proteins causing them to activate, dimerize, and translocate to the nucleus wherein they induce anti-viral gene regulation (Sim et al., 2014). The JAK/STAT pathway has been reported to be utilised by Ae. aegypti to inhibit DENV (Jupatanakul et al., 2017) but not CHIKV (Jupatanakul et al., 2017; McFarlane et al., 2014), whereas, there are contradictory reports for ZIKA (Angleró-Rodríguez et al., 2017; Jupatanakul et al., 2017). Viruses have developed the means to counteract host defence responses. In Ae. albopictus (U4.4) cells, SFV inhibits JAK-STAT signalling (Fragkoudis et al., 2008) and in mammalian cells, SINV and CHIKV assign nsP2 proteins to inhibit INF-induced JAK-STAT signalling independently of transcriptional shutoff (Fros et al., 2010). Clearly, the mechanistic potentiation of JAK-STAT pathways in mosquitoes are unique for different viruses.

In D. melanogaster, pathogen recognition receptors (PRR) activate the signalling cascades Toll, when the pathogen associated molecular patterns (PAMPs) of fungi or gram-positive bacteria are detected or Imd, when the PAMPs of gram-negative bacteria are detected (Fragkoudis et al., 2009).

10 Imd activation can also initiate the JNK pathway (Sluss et al., 1996). Activation of Toll and Imd pathways results in the nuclear translocation of the NF-B/Rel transcription factors Dif or Dorsal for Toll or Relish for Imd. Ae. aegypti orthologs for D. melanogaster transcription factors Dorsal and Relish are Rel1 and Rel2, respectively (Shin et al., 2002; Shin et al., 2005). Rel transcription factors initiate the production of antimicrobial peptides (AMPs) and reactive oxygen species (ROS) (Fragkoudis et al., 2009; Lee et al., 2019). The different AMP families identified in Ae. aegypti include defensins, cecropins, diptericin, attacin and gambicin. Ae. aegypti defensins DefA and DefC are induced following infection with ZIKV and CHIKV and positively regulate antiviral innate immunity (Zhao et al., 2018). Cecropins are upregulated in DENV-2 infected Ae. aegypti mosquitoes and exhibit antiviral activities against both DENV and CHIKV (Luplertlop et al., 2011). In Drosophila, the regulation of Imd and JAK/STAT pathways by attacin C and diptericin B inhibits SINV replication (Huang et al., 2013), although in Ae. aegypti, SINV infection supresses Toll and Imd regulator genes and cecropin, defensin, diptericin and transferrin are significantly downregulated (Kim and Muturi, 2013). In Ae. aegypti salivary glands the Toll pathway genes Gram- negative bacteria-binding protein (GNBP), Toll5A, and myeloid differentiation primary response 88 (MYD88) are upregulated following infection with DENV and silencing MYD88 in the midgut increases DENV replication (Luplertlop et al., 2011). In Drosophila infected with vesicular stomatitis virus (VSV) or Rift Valley Fever virus (RVFV), the activation of the Toll7-induced phosphatidylinositol 3-kinase (PI3K)-Akt-signalling pathway induces autophagy and subsequent viral suppression (Nakamoto et al., 2012; Shelly et al., 2009). Conversely, in Ae. aegypti Aag2 cells, the induction of autophagy by DENV, ZIKV, or CHIKV infection has no discernible effect on viral replication (Brackney et al., 2020). Whereas, the Ae. aegypti protein, venom allergen-1 (AaVA-1), that is produced in the mosquito saliva during infection with DENV or ZIKV and transmitted to a vertebrate host during mosquito feeding, was shown to enhance the viraemia in bitten mice by inducing the autophagy of monocytes (Sun et al., 2020). It has been observed that disparate host species share conserved innate responses to specific pathogens. However, differences in the sensitivities of pathway excitation by even closely related pathogens, such as the induced expression of AMP in Ae. aegypti elicited by the closely related alphaviruses, CHIKV and SINV, exemplifies the nuanced complexities that are involved in potentiating an innate pathway and relative effectiveness in pathogen control that its induction delivers. In mosquitoes, the most effective defence response to viral infection involves the RNA interference (RNAi) pathways.

11 1.8 RNA interference pathways

RNAi was discovered nearly three decades ago by Napoli et al (1990) when their attempts to overexpress chalcone synthase (CHS) in pigmented petunia petals resulted in the CHS gene producing significantly reduced levels of mRNA (Napoli et al., 1990). RNAi has since been recognised as an important mechanism by which eukaryotes regulate transcriptional and post- transcriptional gene expression, mRNA methylation, and chromatin remodelling (Hussain et al., 2016a). The RNAi pathway produces small non-coding RNAs (sncRNAs: 18-30 nt) that are involved in a range of biological processes including host-pathogen interactions and anti-viral responses (Hussain et al., 2016a; Zambon et al., 2006). In insects, the major sncRNAs are short interfering RNA (siRNA), P-element-induced wimpy testis (PIWI)-interacting RNA (piRNA), and microRNA (miRNA) (Figure 1.5).

The siRNA response is initiated when endogenous or exogenous (synthetic or viral) double- stranded RNA (dsRNA) are intercepted and cleaved by Dicer-2 (Dcr2) into 20-22 nt siRNA

12 duplexes (Van Rij et al., 2006) (Figure 1.5). Dcr2 and cofactor R2D2 escort the siRNA duplex onto an Argonaute 2 (Ago2) protein within an RNA-induced silencing complex (RISC) (Ghildiyal and Zamore, 2009). The passenger strand of the siRNA duplex is cleaved from the RISC/Ago2 complex and the remaining guide strand binds with 100% complementarity to mRNA or viral sequences targeted for Ago2 splicing (Ghildiyal and Zamore, 2009). Campbell et al (2008) showed that Ae. aegypti produced complementary siRNAs in response to a SINV challenge and that silencing Ago2 or Dcr2 endonucleases caused a transient increase in SINV replication (Campbell et al., 2008). Cirimotich et al (2009) demonstrated that Flock House virus (FHV: Nodaviridae: Alphanodavirus) B2 protein inhibited the siRNA response by shielding SINV dsRNA from RNAi detection resulting in a fatal viraemia in Ae. aegypti (Cirimotich et al., 2009). Additionally, Ae. aegypti has been shown to use siRNA to control DENV (Franz et al., 2006), CHIKV (McFarlane et al., 2014), and in Ae. aegypti-derived Aag2 cells, SFV (Siu et al., 2011); examples that demonstrate the versatility of the RNAi response.

Retrotransposons are transposable elements (TE) that are bi-directionally transcribed into long single-stranded RNAs that become reverse-transcribed into DNA and reinserted into the genome, often with deleterious effects. Embedded within TE genes are piRNA clusters that, when concurrently transcribed, produce piRNA transcripts that ultimately target and silence TE transcripts. In germ cells, piRNA cluster transcripts are processed into precursor piRNAs (pre-piRNAs) by the endonuclease Zucchini (Zuc) (Nishimasu et al., 2012) (Figure 1.5). Pre-piRNAs are transported to cytoplasmic YB bodies and loaded onto Piwi proteins as part of the RNA induced silencing complex (piRISC) (Murota et al., 2014). Onboard the piRISC complex pre-piRNAs are trimmed to form mature 25-30 nt piRNAs (Nishimasu et al., 2012). Following this, the Piwi-piRNA-RISC complex is transported into the nucleus, wherein, complementary binding of piRNA with TE RNA instigates its cleavage by the endonuclease activity of Piwi (Moazed, 2009). In contrast, secondary piRNAs are generated in a ‘ping pong’ cycle, whereby, pre-piRNAs are loaded onto and processed by Aubergine (Aub) or Ago3 proteins within a piRISC complex that target complementary TE transcripts (Morazzani et al., 2012). Those transcripts consumed by Aub-sense-piRISC provide the template for the formation of Ago3- antisense-piRISC that reciprocally target complementary TE transcripts reproducing Aub-sense- piRISC (Brennecke et al., 2007). Therefore, the biogenesis of secondary piRNAs result from an amplification cycle, whereby, Aub-sense and Ago3-antisense piRNAs play juxtaposed roles in silencing deleterious transposons. Several studies have detected alphavirus-derived piRNAs in mosquitoes infected with SFV, CHIKV, ONNV, and SINV, although the consequence of these piRNA responses remain unclear (Bronkhorst and van Rij, 2014). In Ae. aegypti-derived Aag2 cells, the knockdown of the helicase Spindle-E (SpnE), a protein known to be an imperative co-factor in

13 the piRNA pathway in D. melanogaster, facilitates the increased replication of CHIKV and SFV, however, it was determined that the antiviral activities of SpnE works independently of the siRNA and piRNA pathways (Varjak et al., 2018). In Ae. aegypti, only 19% of total piRNAs are directed against transposons (Arensburger et al., 2011), whereas, a substantial portion of genes involved in germline and embryonic development in An. gambiae encode piRNA transcripts, suggesting that piRNAs in mosquitoes have multiple functions (Hussain et al., 2016a).

In the canonical pathway of miRNA biogenesis primary microRNAs (pri-miRNA) are transcribed from miRNA genes by RNA polymerase II (RNase II) (Bartel, 2009) (Figure 1.5). The single or plural stem-loop structures of pri-miRNAs are cleaved by the RNase III enzyme Drosha in association with Pasha, releasing precursor miRNA (pre-miRNA) (Denli et al., 2004). Pre-miRNAs are transported into the cytoplasm by Exportin-5 (Exp-5) where Dicer-1 (Dcr1) cleaves the pre- miRNA hairpin head producing a 22 nt miRNA duplex comprising 5’ (guide) and 3’ (passenger) strands (Yi et al., 2003). Ago1 or Ago2 proteins load the miRNA duplex onto a RISC complex, where the duplex helix is unwound, and the passenger strand detaches and degrades or is relocated to another RISC complex. Principally, nucleotides 2-8 (seed region) of the remaining miRNA guide strand bind complementary mRNA or viral sequences resulting in translational repression or mRNA degradation (Lewis et al., 2005), however, complementary miRNA-mRNA base pairing outside of the seed region can occur and mRNA binding can cause translational enhancement or mRNA stability (Asgari, 2014). The noncanonical Drosha-independent pathways generate miRNAs through the production of pre- miRNAs during hairpin intron splicing (Okamura et al., 2007), from the cleavage product of tRNAs by tRNase Z (Babiarz et al., 2011), or from small nucleolar RNAs (snoRNA) (Taft et al., 2009).

In mosquitoes, miRNAs are involved in various biological processes including blood digestion, oogenesis, development, and pathogen defence (Hussain and Asgari, 2014). The ingestion of blood triggers a diverse and complex miRNA response (Hussain et al., 2013a). In Ae. aegypti, miR–1174 and miR-1175 are upregulated in the gut (Liu et al., 2014) and in the fat body, miR-8, miR-14, and miR-275 are upregulated and were shown to be involved in digestion and egg development (Bryant et al., 2010). Demonstrating miRNA involvement in pathogen defence is challenging and most studies that assert this link only report a corollary between miRNA differential expression and infection where the involvement of immune-related genes are bioinformatically predicted as putative miRNA targets and rarely validated with definitive experimentation.

14 1.9 Crosstalk between the innate immunity and the RNAi pathways

Maintenance of metabolic balance and homeostasis is coordinated by dynamic fluctuations in global gene expression and miRNAs have been shown to be involved in fine-tuning this balance. Innate immune responses to pathogen invasion involve a cascade of humoral signals activating transcription factors that modulate the transcriptomal profile of immune-related genes, including changes in co-expressed miRNAs. The epigenetic regulation of innate immunity by miRNAs has been demonstrated in mammalian systems (Boosani and Agrawal, 2016), but there are very few insect studies that provide definitive examples of this link. In one such study, Hussain et al. (2013b) conclusively showed in an Ae. aegypti-DENV-2 model that miR-275 is a blood meal-induced immune-retardant that promotes viraemia by targeting cactus and Rel1. More commonly, studies of differential miRNA expression, in the context of infection, are correlative and purports of miRNA- mRNA relationships are usually limited to putative, bioinformatically predicted miRNA targets. In Ae. aegypti infected with DENV-2, Campbell et al. (2014) identified 35 differentially abundant miRNAs that were predicted to be involved in transport, signal transduction, and metabolism and in the saliva, Maharaj et al. (2015) identified five miRNAs that were predicted to be involved in the regulation of CHIKV infection.

Changes in intragenic miRNA expression arise from miRNA genes that are embedded in the exonic regions of a host gene and become coexpressed during transcription (Gulyaeva and Kushlinskiy, 2016). In humans, the dependence of miRNA expression on a host gene was shown to be important in responses to various immune signalling stimuli (Elton et al., 2013). The principal role and utility of miRNAs are to temper or silence mRNA expression and their co-expression with immune-related genes may be a context-specific (e.g. state of infection) mechanism to counterbalance the expression of obsolescent, redundant, or conflicting genes or pathways. Schwenke et al. (2016) showed that during an infection in insects, the balance between the pleiotropic endocrines, juvenile hormone (JH) and 20-hydroxyecdysone (20HE), are modulated by insulin-like growth factor-like signalling (IIS) causing the cessation of egg production (JH↓) (diapause) by reallocating metabolic resources from egg production to immune processes (20HE↑) for defense responses. In Ae. aegypti, the modulation of JH expression during the vitellogenic stage, which is crucial for normal egg production, is tightly regulated by the zinc finger transcription factor Krüppel homologue 1 (Kr‐h1) (Ojani et al., 2018). In An. gambia, it was shown that Let-7 finetunes the expression of Kr-h1 and that marginal differences in Kr-h1 levels above or below optimum can negatively impact oocyte production (Fu et al., 2020). In Drosophila, increased levels of 20HE coincides with the upregulation of Let-7, which targets the AMP diptericin to control its overexpression (Garbuzov and Tatar, 2010).

15 Low IIS activity, which is reciprocally antagonistic to immune signalling, enhances the nuclear localization of the transcription factor FOXO, which can also promote the expression of AMP genes (Becker et al., 2010). It was shown that the upregulation of Drosophila FOXO (dFOXO), which involves innate immunity responses to viral infection, amplifies the RNAi response by regulating siRNA and miRNA pathways (Spellberg and Marr, 2015). In Drosophila, it was shown that miR-277 is involved in Toll-mediated immune responses and is downregulated following infection with the Gram-positive bacteria Micrococcus luteus (Li et al., 2017). In Ae. aegypti, miR-277 fine-tunes the expression of insulin-like peptides 7 (ILP7) and 8 (ILP8) that control the respective deposition and mobilisation of lipids which are an important source of energy for ovarian development. Depletion of miR-277 in Ae. aegypti leads to the dysregulation of ILP7 and ILP8, which causes an excessive accumulation of lipid droplets that is detrimental to ovarian production and egg development. Lipid droplet accumulation coincides with increased expression of FOXO, which is involved in innate immunity pathways and is reciprocally antagonistic to fecundity (Ling et al., 2017). Furthermore, the accumulation of lipid droplets in Ae. aegypti infected with DENV, SINV or the bacteria Enterobacter cloacae was shown to be an antipathogen mechanism associated with the Toll and IMD defence pathways (Barletta et al., 2016). Thus, Let-7 and miR-277 are examples of the conserved roles of miRNA utility that intersect between innate immunity and RNAi pathways through the context- specific (e.g. state of infection) inhibition of reproductive processes in disparate species by interposing effect change in response to bacterial or viral infection. It has been shown that a history of erroneous procedures of miRNA sampling and annotation have caused a spurious accounting of miRNA variability between insect species and that there is greater conservation and less variability than what records imply (Ylla et al., 2016).

1.10 Thesis aims and hypothesis

There are limited studies that demonstrate tissue-specific changes in the miRNA profile of mosquitoes infected with RNA viruses and there are no studies on RRV. Comparative gene expression studies in mosquitoes infected with RRV are prohibitive because none of the important vectors of RRV have been sequenced, which may explain the dearth of studies on this virus. The genome of Culex quinquefasciatus has been sequenced (Arensburger et al., 2010), whereas speculations of this mosquito being associated with RRV transmission are highly unlikely because vector competence studies have shown this mosquito to be poorly susceptible to RRV infection (Kay, 1982). The complete genome of Ae. aegypti has been sequenced and its transcriptome and miRNAome are well characterized. RRV has never been isolated from field-caught Ae. aegypti and this mosquito is not considered to be a vector of RRV, but it has been shown to competently transmit

16 the virus. Further, Ae. aegypti was listed among the possible mosquitoes involved in transmitting RRV during the South Pacific Islands outbreak in 1978-80, and the mosquito has been followed with interest as a potential vector in facilitating the global spread of RRV.

The aims of this study are to investigate the differentially abundant small RNA profiles in the midgut and fat body tissues of Ae. aegypti infected with RRV and to explore the potential role these small RNAs may have in connection with infection. Mosquitoes were fed human blood with or without RRV and collected at 2, 6, and 12 days post feeding. RNA from the respective tissues collected from mosquitoes of each treatment and time point was sequenced and the small RNAs therein were identified. The differential profile of miRNAs, between infected and non-infected samples of the respective tissues and time-points, was analysed and the putative mRNA targets were bioinformatically predicted and compared to topical published literature. Other differentially expressed small RNAs, including siRNA and piRNA, were explored in context with comparable studies on mosquito small RNA responses to alphaviruses. Given that Ae. aegypti has been shown to be readily susceptible to infection with, and competently transmit, RRV, the hypothesis is that the differential small RNA profile of Ae. aegypti to RRV would be comparable to other studies in the literature demonstrating active miRNA and RNAi responses to alphavirus infection.

17 The following publication has been incorporated as Chapter 2.

[1] James Sinclair, and Sassan Asgari, Ross River virus Provokes Differential Expression of MicroRNA and RNA Interference Responses in Aedes aegypti mosquitoes, Viruses, 12, 695, 2020

Contributor Statement of contribution % James Sinclair Writing of text 90 Proof-reading 5 Theoretical derivations 90 Numerical calculations 100 Preparation of figures 95 Initial concept 5 Sassan Asgari Writing of text 5 Proof-reading 95 Supervision, guidance 100 Theoretical derivations 10 Preparation of figures 5 Initial concept 95

18 Chapter 2 RRV provokes Ae. aegypti miRNA and RNAi responses

RRV provokes differentially expressed microRNA and RNAi responses in Ae. aegypti mosquitoes

Abstract: Alphaviruses are globally distributed and predominately transmitted by mosquitoes. Aedes species are common vectors for the clinically important alphaviruses, Chikungunya, Sindbis, and Ross River (RRV) viruses, with Aedes aegypti also being a vector for the flaviviruses dengue, Yellow Fever, and Zika viruses. Ae. aegypti was putatively implicated in the large 1979-80 South Pacific Islands outbreak of RRV; the leading cause of arboviral disease in Australia today. The RNA interference (RNAi) defence response in mosquitoes involves a number of small RNAs with their kinetics induced by alphaviruses poorly understood, particularly at the tissue level. We compared the small RNA profiles between RRV-infected and non-infected Ae. aegypti midgut and fat body tissues at 2, 6, and 12 days post inoculation (dpi). RRV induced an incremental RNAi response yielding short interfering and PIWI-interacting RNAs. Fourteen host microRNAs were differentially expressed due to RRV with the majority in the fat body at 2 dpi. The largely congruent pattern of microRNA regulation with previous reports for alphaviruses and divergence from those for flaviviruses suggests a degree of conservation, whereas patterns of microRNA expression unique to this study provide novel insights into the tissue-specific host-virus attributes of Ae. aegypti responses to this previously unexplored old-world alphavirus.

19 2.1 Introduction

Members of the Alphavirus genus (Togaviridae) are mosquito-borne enveloped RNA viruses that are globally distributed, and those of medical significance are attributed to causing encephalitic [e.g. Eastern (EEEV), Western (WEEV) and Venezuelan Equine Encephalitis (VEEV) viruses] or arthritogenic [e.g. Ross River virus (RRV), Chikungunya virus (CHIKV), and the archetypal Sindbis virus (SINV)] diseases (Contigiani and Diaz, 2017). Alphaviruses are maintained through propagation in sylvatic transmission cycles (animal-mosquito-animal) between amplifying vertebrate hosts in which the titre of viraemia is high and sustained (Weaver and Barrett, 2004). Occasionally, transmission cycles encroach into urban environments via short-term amplifying hosts such as peridomestic animals (e.g. bats, birds, cats, dogs, or possums), however, during epidemics, transmission cycles that are entirely urban (human-mosquito-human) can occur in densely populated regions (Koolhof and Carver, 2017).

RRV is an arthritogenic virus endemic to Australia, Papua New Guinea, and South Pacific Islands (Lau et al., 2017; Russell, 2002b; Togami et al., 2020). The virus causes epidemic polyarthritis (EPA; otherwise known as Ross River disease or fever) and is the leading cause of viral disease in Australia with a rolling annual mean between 2012 and 2017 exceeding 6000 cases (NNDSS, 2018). Based on previous estimates and factoring in inflation, the national cost of EPA in 2018 exceeded AU$9.3 million in medical expenses, whereas the true magnitude of associated costs is unclear when also factoring in mosquito surveillance, control and response activities (e.g. vector control exceeds AU$13 million per annum), or outbreak impact on tourism and industry (Ratnayake, 2006; Tomerini, 2008).

RRV was first isolated from Aedes vigilax near Townsville, Queensland in 1959 (Doherty et al., 1963). In Australia, Ae. vigilax is an epidemiologically important saltmarsh mosquito in the coastal regions of the north [11]. Other important species of mosquitoes include Ae. camptorhynchus in the south (Russell, 2002b), the inland freshwater species Culex annulirostris (Russell, 1994), and Ae. notoscriptus in urbanized regions (Ritchie et al., 1997). Interestingly, Ae. notoscriptus is associated with transmission cycles of RRV involving the Australian brushtail possum (Boyd et al., 2001; Kay et al., 2007), both of which are prevalent in New Zealand (Derraik, 2004; Holder et al., 1999). Although there are no records of RRV circulation in New Zealand, the presence of Ae. notoscriptus and brushtail possums indicates a significant potential for local transmission, if RRV were to be introduced (Kelly‐Hope et al., 2002). There is no definitive record of the species of mosquitoes involved in the Pacific Islands epidemic of 1979-1980, whereas those implicated based on prevalence and capacity to readily transmit RRV included Ae. vigilax, Cx. annulirostris, Ae. 20 polynesiensis, Ae. oceanicus, Cx. quinquefasciatus, and Ae. aegypti (Gubler, 1981; Nasci and Mitchell, 1994). RRV has been isolated from 42 species of mosquitoes (Russell, 1994; Russell, 2002b; Whelan et al., 1993), whereas laboratory competence studies have only indicted 10 as being capable of transmission (Russell, 2002b). RRV has never been isolated from Ae. aegypti in the field and this species of mosquito is not considered a vector, however, multiple competency studies have shown that Ae. aegypti can readily transmit the virus (e.g.(Taylor and Marshall, 1975)).

Ae. aegypti is unrivalled in medical importance being culpable for our most impactful diseases including CHIKV, dengue virus (DENV), Yellow Fever virus (YFV), and Zika virus (ZIKV), wherein, collectively, tens of millions of cases of disease and tens of thousands of deaths occur globally each year (Bhatt et al., 2013; Musso and Gubler, 2015; Powell, 2018; Tabachnick, 1991). The medical importance of Ae. aegypti is underscored by its capacity to adapt to a range of ecological niches, both spatial and temporal, to transmit multiple disparately related and highly pathogenic viruses, and for its natural drive to preferentially seek out human blood. The capacity of Ae. aegypti to vector a wide range of viruses and its ability to adapt to foreign environments stresses our need to better understand the molecular host-virus interactions in this mosquito.

Genomic studies analysing differentially regulated genes has enabled significant progress in elucidating mosquito innate immune responses to arbovirus infection. Sim et al (2005) pioneered this approach reporting the modulation of Anopheles gambiae gene expression in response to O’nyong- nyong virus (ONNV: Alphavirus) infection, and later in 2007 they demonstrated that this modulation resulted in the production of heat shock protein cognate 70B (HSC70B), which impeded ONNV replication (Sim et al., 2005; Sim et al., 2007). In response to arboviral infection, Ae. aegypti has been reported to initiate Toll (Xi et al., 2008), c-jun N-terminal kinase (JNK) (Sanders et al., 2005), immune deficiency (Imd) (Luplertlop et al., 2011), and Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathways (McFarlane et al., 2014). The JAK-STAT pathway has been reported to be utilized by Ae. aegypti to inhibit DENV (Jupatanakul et al., 2017) but not CHIKV (Jupatanakul et al., 2017; McFarlane et al., 2014), whereas there are contradictory reports for ZIKV (Angleró-Rodríguez et al., 2017; Jupatanakul et al., 2017).

RNA interference (RNAi) has been recognized as an important mechanism by which eukaryotes regulate transcriptional and post-transcriptional gene expression, mRNA methylation, and chromatin remodelling (Hussain et al., 2016a). The RNAi pathway produces small non-coding RNAs (sncRNAs: 20-30 nt) that are involved in a range of biological processes including host-pathogen interactions and anti-viral responses (Blair and Olson, 2015; Hussain et al., 2016a). In insects, the

21 major sncRNAs are short interfering RNA (siRNA), P-element-induced-wimpy-testis (PIWI)- interacting RNA (piRNA), and microRNA (miRNA).

Campbell et al (2008) showed that Ae. aegypti produced complementary siRNAs in response to a SINV challenge and that silencing Argonaut 2 (Ago2) or Dicer-2 (Dcr2) endonucleases caused a transient increase in SINV replication (Campbell et al., 2008). Concomitantly, Cirimotich et al (2009) demonstrated that Flock House virus (FHV: Nodaviridae, genus Alphanodavirus) B2 protein inhibited the siRNA response by shielding SINV double stranded RNA (dsRNA) from RNAi detection resulting in a fatal viraemia in Ae. aegypti (Cirimotich et al., 2009). Additionally, Ae. aegypti has been shown to use siRNA to control DENV (Franz et al., 2006), CHIKV (McFarlane et al., 2014), and in Ae. aegypti-derived Aag2 cells, Semliki Forest virus (SFV: Togaviridae, genus Alphavirus) (Siu et al., 2011); examples that underscore the versatility of the RNAi response.

Several studies have detected alphavirus-derived piRNAs in mosquitoes infected with SFV, CHIKV, ONNV, and SINV, although the consequence of these piRNA responses remain unclear (Bronkhorst and van Rij, 2014). In Ae. aegypti, only 19% of total piRNAs are directed against transposons (Arensburger et al., 2011), whereas a substantial portion of genes involved in germline and embryonic development in An. gambiae encode piRNA transcripts, suggesting that piRNAs in mosquitoes have multiple functions (Hussain et al., 2016a).

The ingestion of viraemic blood by a mosquito results in a diverse and complex miRNA response (Hussain et al., 2013a). Liu et al (2015) showed that miR–1174 and miR-1175 were upregulated in the gut of Ae. aegypti following a blood meal (Liu et al., 2014). Bryant et al (2010) showed that miR-8, miR-14, and miR-275 were upregulated in the fat body of Ae. aegypti following a blood meal and that these are involved in digestion and egg development (Bryant et al., 2010). Campbell et al (2014) reported 35 differentially abundant miRNAs in Ae. aegypti in response to DENV2 infection with the majority of predicted targets involved in transport, signal transduction, and metabolism (Campbell et al., 2014). In the saliva of Ae. aegypti infected with CHIKV, Maharaj et al (2015) identified five miRNAs that were predicted to be involved in the regulation of CHIKV infection (Maharaj et al., 2015).

There are very few studies demonstrating miRNA responses in mosquitoes infected with RNA viruses at the tissue level. In particular, there are no studies on RRV. This may be because none of the important vectors of RRV have been sequenced and therefore a platform for comparative genomics to an epidemiologically relevant reference cannot be made. The genome of Culex Quinquefasciatus has been sequenced (Arensburger et al., 2010), although vector competence studies

22 have shown this mosquito to be poorly susceptible to RRV infection and, therefore highly unlikely to be involved in RRV maintenance (Kay, 1982). Ae. aegypti has been shown to competently transmit and potentially facilitate the global spread of a range of viruses, including RRV, which stresses our need to better understand the repertoire of molecular host-virus interactions in this mosquito. Further, the complete genome, transcriptome and miRNAome of Ae. aegypti are well characterized. In this study, we reveal the differentially abundant small RNA profile of the pertinent organs (midgut and fat body) in Ae. aegypti infected with RRV. These results provide an important and relevant contribution to our understanding of the molecular host-virus interactions within a medically significant vector infected with an old-world alphavirus.

2.2 Materials and Methods 2.2.1 Ethics statement

The T48 strain of RRV used in this study was kindly provided by Jody Peters of the School of Chemistry and Molecular Biosciences, The University of Queensland. The virus was originally isolated from Ae. vigilax mosquitoes (Townsville, QLD, Australia). RRV protocols were approved by the University of Queensland Biological Sciences Biosafety Committee (Reference number: IBC/231B/SBS approved on 22 May 2018).

2.2.2 Mosquito infections with Ross River virus

Ae. aegypti (Innisfail strain) mosquitoes were propagated from eggs and raised in an insectary maintained at 27C, 80% relative humidity (RH), and a 12 h photoperiod. RRV (T48 strain) was amplified in C6/36 cells (Ae. albopictus) cultured at 28C in a 1:1 v/v mixture of Schneider's Drosophila Medium (Life Technologies, California, USA) and Mitsuhashi and Maramorosch (Himedia, Mumbai, India) medium supplemented with 2% foetal bovine serum (FBS). The virus was then amplified in African green monkey kidney (Vero) cells cultured at 37C in Opti- MEM (Life Technologies, California, USA) media with 2% FBS for 48 h. Control C6/36 and Vero cells that did not include virus were cultured each time and in the same way. The same volume of C6/36 cell supernatant that was used to inoculate Vero cells was also transferred between control cultures. Virus amplified in Vero cells was mixed 1:1 v/v with whole human blood (Australian Red Cross) and 0.1 % ATP (phagostimulant). Females aged 4-6 days old were fed blood mixtures with or without virus at 2 × 105 plaque forming units per mL (pfu/mL). At 2, 6, and 12 days post-feeding (dpf) mosquitoes were cold anaesthetized and transferred to −80C storage until use.

23 2.2.3 Mosquito dissection and RNA extraction

The midgut and fat body tissues were dissected from a total of 180 mosquitoes yielding 360 organs. The organs were pooled in lots of 10 yielding 36 samples comprising the tissues fat body (N = 18) or midgut (N = 18) that were infected (N = 9) or uninfected (N = 9) at day 2 (N = 3), day 6 (N = 3), or day 12 (N = 3). Total RNA was extracted from samples (N = 36) using the miRNeasy Mini Kit following the manufacturer’s protocol (Qiagen, Hilden, Germany). For practical reasons, we did not confirm infection or non-infection in all mosquitoes (N = 180). During dissection, the RNA from the carcass remnants of 10 randomly selected mosquitoes sampled from each day and treatment was used to detect virus presence or absence (control) by RT-qPCR on a Rotor-Gene Q (Qiagen, Hilden, Germany) machine using ribosomal protein subunit 17 (RPS17) and in-house-designed RRV-specific primers (Table S1). Mean normalized expression (MNE) of samples relative to RPS17 was calculated using the second derivative maximum take-off and amplification efficiency values produced by the Rotor-Gene Q software.

2.2.4 Library preparations and sequencing

The TruSeq Small RNA protocol (Illumina) was used to generate and sequence complementary DNA (cDNA) libraries from small-input RNA on a short-read sequencing platform with an Illumina Hi-Seq machine (Illumina). Briefly, cDNA libraries were amplified from 1 µg of size-fractionated total RNA ligated with indexed 3’ and 5’ adapters (Table S1). Libraries were size-selected by polyacrylamide gel, purified, and concentrated. An Agilent Technologies 2100 Bioanalyzer (California, USA) was used to assess purity for quality control. Illumina RTA software (1.17.21.3) was used for base-calling for quality control and Illumina bcl2fastq 2.17 software was used for demultiplexing. Final raw reads were in FASTQ format with the 3’ adapter retained. The raw data for infected and uninfected Ae. aegypti samples generated in this study have been deposited in the Sequence Read Archive (SRA) under the accession PRJNA635740.

2.2.5 Small RNA analysis

The software tool Fastq_quality_filter (FASTX-toolkit) (v0.0.6) (Pearson et al., 1997) was used to earmark raw reads that had at least 95% of their sequence content with a phred quality score above 20. Using Flexbar (Dodt et al., 2012), we trimmed the 3’ adaptors (Table S1), allowing a minimum overlap of 4 nucleotides (nt), and remaining reads with less than 16 nt were discarded. The length distribution of reads was visualized using scales (Wickham and Seidel, 2019), plyr (Wickham, 2011) and ggplot2 (Wickham, 2016) in the R package (v3.6). The miRPro pipeline (Shi et al., 2015) was

24 used to enumerate and categorize the samples as miRNA content and other RNA biotypes. Cleaned reads were mapped to the most recent miRBase repository (v21) (Griffiths-Jones et al., 2006) using Novoalign (Novocraft, Selangor, Malaysia) and both mature and novel mature miRNAs were enumerated with htseq-count (HTSEQ 1.11) (Anders et al., 2015) and Samtools algorithm (Li et al., 2009). Remaining reads were mapped to the most current Ae. aegypti genome (AaegL5; VectorBase) (Giraldo-Calderón et al., 2015) for RNA reads categorization.

Differential expression of miRNAs (Table S2) was calculated from concatenated mature and novel mature miRNA counts using DESeq2 (v3.11) (Love et al., 2014) following various namesake Bioconductor vignettes in R (Huber et al., 2015; Team, 2013). The counts from all samples were included in the experimental design for differential expression which included the three main factors (1) tissue (tsu), (2) time (day), and (3) condition (cnd) and interaction terms between these factors (design = tsu + tsu:day + tsu:day:cnd). The levels for each factor were (1) tsu: fat body (fb) and midgut (mg), (2) day: 2, 6, and 12, and (3) cnd: control (con) and infected (rrv). Fat body, day 2, and control was set as the reference levels by which change was compared. The coefficients derived from the design enabled comparison between tissues mg and fb irrespective of infection (tsu_mg_vs_fb), the effect of age between days with reference to deviance from day 2 (reference) irrespective of infection (e.g. tsu.fb.day6 or tsu.mg.day12), the effect size of condition within tissues within day (e.g. tsufb.day2.cndrrv), within tissue between days (e.g. contrast = tsufb.day2.cndrrv and tsufb.day6.cndrrv), or between tissues within day (e.g. contrast = tsufb.day2.cndrrv and tsumg.day2.cndrrv). The effect size of condition between tissues between days (e.g. tsufb.day2.cndrrv and tsumg.day12.cndrrv) was not explored.

To test for significance, we used the Wald test (alpha, 0.05) for the pairwise statistical comparison of probability calculated between the coefficients generated by our experimental design. The threshold for significance was set at a false discovery rate (FDR) of 0.05. The FDR implemented in DESeq2 is an adjusted p-value given by the Benjamini and Hochberg algorithm (Benjamini and Hochberg, 1995). We used the approximate posterior estimation for generalized linear model (apeglm) function (Zhu et al., 2019) to exclude results from low or highly variable counts and further excluded results with a baseMean < 100. In DESeq2, the term baseMean refers to the mean normalized count across all samples. Remaining miRNAs were considered significantly differentially expressed (SDE). In summary, our criteria qualifying miRNA significance were for a normalized mean count of samples higher than 100, a FDR less than 0.05, and an effect size fold change (FC) of 1.5.

25 2.2.6 RNAi activity analysis

For siRNA and piRNA identification, reads that did not align with either the Ae. aegypti genome (AaegL5) or miRBase records were mapped to the genome of the T48 strain of RRV (GenBank: GQ433359.1) (Jones et al., 2010) using Novoalign (Novocraft, Selangor, Malaysia). BAM alignment files were filtered by strand and lengths of 21nt or 27-29 nt for siRNA and piRNA reads analyses, respectively. We used kpPlotBAMCoverage in karyoploteR (v1.14.0) within the R package (v3.6) (Gel and Serra, 2017) to visualize the genome coverage of siRNA and piRNA reads. We used

WebLogo software (v3) (Crooks et al., 2004) to visualize U1 and A10 piRNA signatures, and 10 nt corroborative overlap signatures were assessed and scored using the public Galaxy server at https://mississippi.sorbonne-universite.fr (Afgan et al., 2018).

2.2.7 miRNA target identification

The algorithms miRanda (Enright et al., 2003), RNAhybrid (Krüger and Rehmsmeier, 2006), TargetScan (Agarwal et al., 2018), PITA (Kertesz et al., 2007), and RNA22 (Loher and Rigoutsos, 2012) were used to predict miRNA targets against Ae. aegypti 3’UTR sequences. For this, a bed file of headers containing the phrase three_prime_utr and corresponding coordinates was extracted from an annotation file (Aedes-aegypti-LVP_AGWG_BASEFEATURES_AaegL5.2.gtf) with grep and used as a template for BEDTools (Quinlan and Hall, 2010) to extract those corresponding 3’UTR sequences from file (Aedes-aegypti-LVP_AGWG_CHROMOSOMES_AaegL5.fa). Sequences shorter than 30 nt were discarded, and to remove duplicates for genes with splice variants, we used only the longest 3’UTR. All target prediction programs were operated locally on a Linux machine using default command line parameters. We referred to the results exclusion criteria used by Oliveira et al (2017) to guide the exclusion criteria we used in this study. For miRanda, we used a minimum score of 150 and a minimum folding energy (mfe) of < −18 (kcal/mol) (Oliveira et al., 2017). For RNA22, we used a p-value of < 0.1 and mfe values < −12 (kcal/mol). For RNAhybrid, we set the p- value at 0.05 and mfe values < −20 (kcal/mol). For PITA, we used mfe values < −18 and excluded 6mer types from both PITA and TargetScan predictions. The intersect of targets predicted by at least three programs were used for downstream analysis (Table S3).

2.2.8 Gene Ontology

The Module Functional Analysis in OmicsBox (Biobam, Valencia, Spain) was used for gene ontology (GO) annotation and analysis (Table S4) of the miRNA targets predicted by at least three programs (Götz et al., 2008). This module used National Centre for Biotechnology (NCBI) Blast

26 against homologous Ae. aegypti sequences to establish the protein identity of the query sequences from this study. The corresponding GO IDs and GO names of the mapped and annotated proteins were retrieved using OmicsBox InterPro scan and these were linked with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Table S5). The statistics and graphical representation of all the results was produced by the OmicsBox program.

2.2.9 RT-qPCR analysis

The miScript II RT kit (Qiagen, Hilden, Germany) was used following the manufacturer’s protocol to synthesize cDNA from 500 ng RNA per 20 µL reaction using the 5× HiSpec Buffer and the proprietary-sequence universal primer on a Veriti thermal cycler (Applied Biosystems, California, USA). cDNA samples were diluted in 200 µL nuclease-free water and 2 µL thereafter amplified in 10 µL duplicate reactions with miRNA-specific primers (Integrated DNA Technologies, Iowa, USA) by RT-qPCR using miScript SYBR Green PCR kit reagents (Qiagen, Hilden, Germany) on a 72-well Rotor-Gene Q thermocycler (Qiagen). The thermocycling parameters used for amplification were 95C for 15 min, followed by 40 cycles of 94C for 15 s, 60C for 30 s, and 70C for 30 s. The mean normalized expression (MNE) of samples was calculated using the second derivative maximum and corresponding amplification efficiencies generated by the Rotor-Gene Q software (Pfaffl, 2001). The

Ae. aegypti small nuclear RNA U6 was used as the calibrator. The log2 ratio between control and treatment MNE values were used to calculate fold change. Three biological replicates were used per treatment.

2.3 Results and Discussion 2.3.1 Illumina sequencing of small RNAs

An Illumina sequencing platform was used to produce small RNA profiles of RRV-infected and non-infected Ae. aegypti mosquitoes. To investigate host miRNA and RNAi responses to RRV infection, RNA samples were extracted from midgut and fat body tissues of mosquitoes collected at 2, 6, and 12 days post-infection (dpi). The rates of viral infection were determined from a subset of individual mosquitoes (N = 10) from days 2, 6, and 12 and the relative viral load in those mosquitoes with a detectable degree of RRV infection (N = 6) was confirmed by RT-qPCR, which demonstrated a progressive increase in viral load over time (Figure 2.1). No virus was detected in uninfected negative control mosquitoes. Considering that carcasses were used for determination of virus infection after dissection of tissues, 100% dissemination of virus at 12 dpi suggested high rate of infection even if it was detected in about 50% of mosquito carcasses at 2 dpi when virus is at early stages of infection in the midgut, which was already removed from mosquito bodies.

The combined yield of raw data obtained from control and infected small RNA libraries was 200 and 232 million reads, respectively (Table S2). Of the reads, we discarded 6-28% in different libraries due to their low-quality score or lack of adapter sequence. A total of 94% of the annotated Ae. aegypti miRNAs referenced by miRBase were found among all the samples in our data representing 7-22% of clean reads in different libraries. The sequence length distributions in all libraries expressed peaks at 21-23 nt, consistent with the characteristic lengths of miRNA and short interfering RNA (siRNA), and peaks at 27-29 nt representing piRNA - a common feature of insect small RNA libraries (Figure 2.2).

2.3.2 Differential expression of Ae. aegypti miRNAs in response to RRV infection

Small RNA library analysis of RRV-infected Ae. aegypti fat body and midgut tissues identified 14 SDE miRNAs at different time points compared with uninfected mosquitoes (Figure 2.3). Interestingly, miRNA modulation was prominent early in infection at 2 dpi, and by day 12, when the viral load was higher, most miRNAs in both tissues were stably expressed (Figure 2.4). In contrast to siRNAs, there is very little evidence that miRNAs directly target viral genomes (Trobaugh and Klimstra, 2017), and therefore we would not expect to see the positive correlation of changes in miRNA expression with viral load. It is then perhaps not surprising to see the dramatic shift in miRNA expression during the early adjustment phase to infection when the greatest number of changes in target gene expression occur. The identified miRNAs were predominately downregulated (60%) and from the fat body (93%) tissue (Table 1). An intriguing observation is the preponderance for the downregulation of Ae. aegypti miRNAs reported for both flavivirus (DENV and ZIKV) (Campbell et al., 2014; Saldaña et al., 2017) and alphavirus (CHIKV and SINV) (Maharaj et al., 2015; Myles et al., 2008; Shrinet et al., 2014) infections; although direct comparison between these studies and our results is difficult considering different time points or samples used (whole mosquito/tissues/cell line). Myles et al (2008) reported an overall depletion of miRNA expression in Ae. aegypti infected

29 with SINV, which was the first record in the literature investigating the miRNA expression of Ae. aegypti infected with an old-world alphavirus, when Ae. aegypti miRNAs had not yet been annotated and the authors’ analysis for differential expression used Drosophila melanogaster miRNAs as reference (Myles et al., 2008).

Figure 2.4 RRV caused differential expression of Aedes aegypti miRNAs mostly early in infection. Volcano plots show miRNA log2 fold changes due to RRV infection in fat body at (A) 2 dpi, (B) 6 dpi, and (C) 12 dpi, and in midgut at (D) 30 2 dpi, (E) 6 dpi, and (F) 12 dpi. Significantly differentially expressed miRNAs (orange) were found in fat body samples at 2 dpi and 12 dpi and in midgut samples at 2 dpi.

Table 2.1 Differentially expressed Aedes aegypti miRNAs upon RRV infection. Tissue 1 dpi miRNA Control 2 RRV 2 FC 3 p-value FDR FB 2 miR-9b-5p 4752 10939 2.30 9.10 × 10-10 9.55 × 10-8 FB 2 miR-317-3p 22084 13362 -1.65 4.81 × 10-6 2.52 × 10-4 FB 2 miR-275-3p 12465 6290 -1.98 3.74 × 10-5 1.31 × 10-3 FB 2 miR-275-5p 2208 752 -2.94 4.39 × 10-4 9.22 × 10-3 FB 2 miR-8-5p 9926 5071 -1.96 2.41 × 10-4 6.32 × 10-3 FB 2 miR-2945-3p 9070 14292 1.58 1.40 × 10-3 0.02 FB 2 miR-9a-5p 4779 8276 1.73 1.56 × 10-3 0.02 FB 2 miR-988-3p 1340 2062 -1.53 1.68 × 10-3 0.02 FB 2 miR-34-3p 852 523 -1.63 1.69 × 10-3 0.02 FB 2 miR-184-3p 149602 93549 -1.60 1.83 × 10-3 0.02 FB 2 miR-2940-5p 29034 16911 -1.72 1.89 × 10-3 0.02 FB 2 miR-989-3p 11468 67138 5.85 2.18 × 10-3 0.02 FB 12 miR-71-3p 498 366 -2.01 3.85 × 10-5 9.58 × 10-3 FB 12 miR-2940-3p 11495 22203 1.93 4.19 × 10-4 0.05 MG 2 miR-9b-5p 9130 15236 1.66 1.67 × 10-4 0.02 1 FB, fat body. 1 MG, midgut. 2 Base mean counts. 3 FC, fold change.

miR-9b-5p was upregulated in the midgut (FC 1.67) at 2 dpi and was the only SDE miRNA found in midgut samples (Figure 5A). This miRNA was also upregulated in the fat body at 2 dpi (FC 2.29) which was the most significant result from this study (FDR, 9.55 × 10-8). Maharaj et al (2015) reported downregulation of this miRNA from a read count of 64 to 34 in the saliva of Ae. aegypti infected with CHIKV at 10 dpi, although the change in miR-9b expression was not considered statistically significant (Maharaj et al., 2015). By contrast, in Ae. albopictus, Liu et al (2015) found miR-9b to be significantly downregulated at 7 dpi with DENV. Considering about 50% virus dissemination was detectable in mosquitoes at 2 dpi (Figure 1A), we speculate that the early changes in fat body miRNAs could be part of an early systemic response to virus infection following the receipt of cytokines from infected midgut cells via hemolymph. Crosstalk between tissues for immune induction following blood feeding has been documented (reviewed in (Pakpour et al., 2014)). In contrast to the significant downregulation of miR-2940-3p in the study by Liu et al (2015), in our study, miR-2940-3p was upregulated. By comparison, miR-184-3p and miR-2940-5p were significantly downregulated in Liu et al (2015) (Liu et al., 2015) and these miRNAs were also downregulated in our study. The continued enquiry of miRNA profile changes elicited by disparate pathogens in this mosquito or other dipterans is required to better clarify their evolutionary conservation and functional significance in response to infection.

The most abundant SDE miRNAs found in this study−miR-184-3p, miR-275-5p, miR-2940- 5p, miR-317-3p, and miR-989-3p−are commonly reported as highly abundant and widely conserved

31 (Etebari and Asgari, 2014). Additionally, miR-989-3p (FC 5.85) and miR-275-5p (FC −2.94), which were both from fat body tissue at 2 dpi, had the most extreme fold changes. By comparison, in An. gambiae infected with Plasmodium bergei, miR-989 was also significantly upregulated (FC 4) and deemed exclusive to midgut tissue (Winter et al., 2007), whereas other studies had contended that, with reference to results comparing the conserved homologues in both Anopheles stephensi and Ae. aegypti, miR-989-3p is predominantly expressed in ovaries (Akbari et al., 2013), and rarely in the midgut tissue (Mead and Tu, 2008). We detected miR-989-3p in the midgut tissue at 2 dpi, whereas its change in expression compared to uninfected samples was not statistically significant. Other studies reporting the significant differential expression of miR-989-3p involved the vectors Ae. albopictus and Culex quinquefasciatus infected with West Nile virus (WNV) (Skalsky et al., 2010), Ae. aegypti infected with DENV (Etebari et al., 2015) or ZIKV (Saldaña et al., 2017), or Ae. aegypti cells (Aag2) infected with Wolbachia (Mayoral et al., 2014). According to our criteria for significance, there were no SDE miRNAs detected from mosquitoes at 6 dpi in either tissue.

The only SDE miRNAs from day 12 mosquitoes were miR-71-3p and miR-2940-3p in the fat body tissue. An analysis of the change in expression of miR-71-3p revealed that in control mosquitoes this miRNA became progressively more abundant throughout the time course, whereas in the fat body at 12 dpi miR-71-3p was significantly downregulated. By comparing samples across time, it appeared that the condition effect due to RRV was an impediment to the natural age-related upregulation of miR-71-3p by day 12 (Figure 5B). It would be interesting to investigate the target genes of miR-71- 3p, the consequence of their not being silenced at that age, and whether their continued expression was part of a deliberate anti-viral defence response or caused by the virus.

Saldaña et al (2017) also noted a disproportionately low number of SDE miRNAs in Ae. aegypti mosquitoes infected with ZIKV, despite the pronounced viral load reached by day 14 (Saldaña et al., 2017). Comparing results for miR-2940-3p from this study and Saldaña et al (2017), the miRNA shows remarkably similar measures of abundance and degree of upregulation (FC ~ 2). While whole Ae. aegypti mosquitoes were infected with a flavivirus and expression had peaked earlier at day 7 in Saldaña et al (2017), the measures of high abundance are consistent with published results for this miRNA from both flaviviruses (Liu et al., 2015; Saldaña et al., 2017) and alphaviruses (Shrinet et al., 2014). Wolbachia is an endosymbiotic bacterium that is currently used to control DENV spread as its presence in Ae. aegypti blocks replication of a number of viruses (Hedges et al., 2008). In Ae. aegypti, Wolbachia induces and utilizes miR-2940 to suppress the expression of Pelo protein (Asad et al., 2018), upregulate host metalloprotease f41 ftsh transcripts (Mayoral et al., 2014), and downregulate the expression of Ae. aegypti DNA methyltransferase 2 (Zhang et al., 2013), activities that aid its preservation in the host and contribute to inhibition of replication and subsequent transmission of co- infected DENV.

Mosquito miRNA induction by CHIKV has been studied extensively, however, research into other alphaviruses is needed to supplement the literature for comparing between species to further develop a characterization of their distinct molecular responses and elucidate with greater clarity the

33 functional reasons for miRNA involvement and why conservation or divergence in miRNA form and function has evolved. Because this project is a foray into the Ae. aegypti miRNA responses to an old- world alphavirus, we prioritized our analysis of miRNA function to those also published with reference to CHIKV or Ae. aegypti fat body.

2.3.3 RT-qPCR validation of differentially expressed miRNAs

The relative expression of 10 miRNAs was ascertained by RT-qPCR from RNA samples extracted from fat body and midgut tissues from control and RRV-infected samples at 2 dpi to validate the RNA-Seq data and subsequent assessment of differential expression of miRNAs (Figure 6). The 10 miRNA candidates for this experiment were selected to represent the scope of fold changes determined by DESeq2, which included those showing the most (miR-989-3p) and least (miR-999- 3p) divergence. The direction of change was mostly consistent between methods with discrepancies for miR-1-3p and miR-2940-3p from midgut and miR-999-3p, bantam-5p, and miR-989-3p from fat body, which notably also showed the largest discrepancy between the reported magnitudes of fold change.

34 2.3.4 miRNA target analysis

The putative Ae. aegypti 3’UTR mRNA targets of the SDE miRNAs determined in this study (Table 2.1) were predicted using the default parameters of miRanda, RNAhybrid, TargetScan, RNA22 and PITA software. From the intersect of targets predicted by at least three programs, we found 528 genes as putative binding sites for at least one miRNA. Among those, 65 genes were associated with at least 2 miRNAs, 6 genes with 3 miRNAs, and AAEL003953 (LOC5579800, ec:2.7.1.123) was found to be a putative target for miR-9b-5p, miR-275-5p, miR-2940-3p, and miR- 2940-5p (Figure 2.7). The greatest number of putative binding sites were for miR-989-3p (148), miR- 9b-5p (98), and miR-2940-3p (53), which were also determined in this study to have the highest upregulation and by the same rank (Table 2.1).

Figure 2.7 Venn diagram showing the number of predicted targets for each miRNA and the intersect shows the number of predicted targets in common.

35 An analysis of the KEGG pathways revealed a number of binding sites that were for genes putatively involved in mosquito immunity and signalling (Table S5). There were 328 sequences with the molecular function annotation (GO:0003674) comprising genes ascribed to hydrolase (78), polymerase (82) and iron binding (57) activities (Figures 8 and 9). Both miR-989-3p and miR-9b-5p were associated with genes involved in zinc-finger protein and the MAPK signalling pathway as described by the response to stress ontology (GO:0006350). It has been shown that the expression of zinc-finger antiviral protein inhibits the replication of members of the Alphavirus genus including SINV, SFV, VEEV, and RRV (Bick et al., 2003). AAEL013596 is a phosphatidylinositol 3-kinase (PI3K) regulatory subunit and involved in the Toll pathway. In Drosophila, the Toll-7 transmembrane receptor activates antiviral autophagy facilitated by the PI3K-Akt-TOR signalling pathway (Nakamoto et al., 2012), whereas for SINV, which forms membrane bound replication complexes, the activation of this pathway in Drosophila becomes pro-viral facilitating its cap-dependent viral RNA translation (Patel and Hardy, 2012). In human cells, the metabolic changes induced by the hyperactivation of PI3K-Akt by SFV and RRV is also pro-viral (Mazzon et al., 2018).

The beta subunit kinase, AAEL003245, is an inhibitor of nuclear factor kappa B and involved in the mTOR signalling pathway, which relates to vitellogenesis, protein synthesis and motility. Genes were associated with the Wingless signalling pathway (Wnt) which works synergistically with the mTOR pathway in regulating vitellogenesis (Gulia-Nuss et al., 2011). Following a blood meal, miR-8 is significantly upregulated in Ae. aegypti fat body and was shown to target the Wnt pathway to promote the production of yolk protein precursors (YPP) and lipid accumulation used for

36 vitellogenesis (Lucas et al., 2015). Conversely, in this study miR-8 was significantly downregulated in the fat body of RRV-infected Ae. aegypti mosquitoes which may signify an active response to stall vitellogenesis. Schwenke et al (2016) showed that in female insects pleiotropic signalling mechanisms regulate the allocation of metabolic resources between reproductive and immune processes and that during infection egg production is stopped to reallocate energy resources to the defence response (Schwenke et al., 2016). In humans, elements of the mTOR signalling pathway are modulated by Influenza A virus (Kuss-Duerkop et al., 2017) and ZIKV (Liang et al., 2016), which promotes viral replication during late stage cell stress.

Figure 2.9 Molecular function. The number and percentage of genes from this study ascribed to various molecular functions.

To corroberate our findings with published evidence, we compared our list of predicted miRNA target genes with studies involving changes in the Ae. aegypti transcriptome due to arbovirus infection. We only included transcripts that were reported as significantly modulated and predicted as a target for at least one miRNA by no less than three algorithms. In total, we found 18 unique genes, representing the predicted targets for 12 out of the 14 miRNAs from our study, which were shown to be significantly modulated in different studies involving whole Ae. aeygpti, midgut and carcass with CHIKV, DENV, WNV, or ZIKV at 1 dpi, 2 dpi, 4 dpi, 7 dpi, or 14 dpi (Dong et al., 2017b; Etebari et al., 2017; Raquin et al., 2017; Zhao et al., 2019). We focused on significantly downregulated transcripts that were predicted as targets for the significantly upregulated miRNAs in this study. Although no genes that was predicted by all five algorithms were found, our most significantly upregulated miRNAs, miR-9b-5p and miR-989-5p, which had most of their targets

37 predicted by four algorithms, were strongly represented (Table S6). Among the upregulated miRNAs, 80% were paired with downregulated transcripts but only 22% of the downregulated miRNAs were paired with upregulated transcripts. The top ten most downregulated genes included studies of DENV, WNV, and ZIKV and were from abundances measured early during the course of infection predominately from 1 and 2 dpi. The greatest number of pairings was from transcript abundances measured early during the course of infection at 1, 2, and 4 dpi.

2.3.5 RRV is a target of the Ae. aegypti RNAi response

The dsRNA intermediates produced during viral replication become a target of the mosquito RNAi defence response, as reported for a number of alphaviruses including SINV (Campbell et al., 2008; Khoo et al., 2010; Myles et al., 2008), SFV (Siu et al., 2011), and CHIKV (McFarlane et al., 2014). The ribonuclease Dicer-2 cleaves these intermediates into 21 nt double stranded short interfering RNAs (siRNAs) which when bound to the RNA-induced silencing complex (RISC) become a template of the viral RNA sequence and guide for subsequent cleavage. To investigate potential RNAi activity against RRV, we mapped reads to the virus genome that had not aligned with either the Ae. aegypti genome or miRBase records. There were 30,652, 238,857, and 420,187 reads in total from days 2, 6, and 12, respectively, which mapped to the viral genome with a length distribution of 15-51 nt. The ratio between reads from fat body and midgut were 1:1.67, 1:0.56, and 1:1.26 for days 2, 6, and 12, respectively. The production of 21 nt sequences, a hallmark of an siRNA response to infection, was evident from day 2, and these increased throughout the time course for both tissues (Figure 2.10A). The distribution of reads mapped somewhat evenly across the genome with a greater number in each case mapping to the sense strand (Figure 2.10B). These results confirm an active RNAi response of Ae. aegypti to RRV infection.

2.3.6 Production of RRV-derived vpiRNAs

In Ae. aegypti, there are eight members of the PIWI protein clade (Piwi 1-7 and Ago3) that are involved in piRNA production from transposons, mRNA or viral RNA. The Tudor protein Veneno recruits Yb, Piwi5 and Ago3 (Joosten et al., 2019) to mediate virus-derived piRNA (vpiRNA) production (Miesen et al., 2015) by ping-pong amplification (Vodovar et al., 2012), where sense and anti-sense vpiRNA pairs show a nucleotide bias for uridine at position one (U1) and arginine at position ten (A10), respectively (Brennecke et al., 2007; Gunawardane et al., 2007). Alphavirus- derived piRNAs have been detected in mosquitoes infected with SFV, CHIKV, ONNV, and SINV (Bronkhorst and van Rij, 2014).

39 To investigate if vpiRNAs are produced in Ae. aegypti infected with RRV, we mapped 27-29 nt reads from 2, 6, and 12 dpi to the RRV genome. The number of reads that mapped to the viral genome increased throughout the time course as the infection progressed with a bias for reads mapping to the sense strand (Figure 2.11). At 6 and 12 dpi for both fat body and midgut samples, there was a moderate clustering of reads at the 3’ end of the genome, which encodes the subgenomic RNA for the production of non-structural proteins (Figure 2.11). The subgenomic promoters of the alphaviruses SINV, CHIKV, and SFV have been shown to drive subgenomic RNA production in mosquitoes, and in those studies, reads mapped in distinct hotspots to the 3’ non-structural end of the viral genome spanning the subgenomic RNA, and almost exclusively to the sense strand (Miesen et al., 2016). In our study, clusters of reads were also formed in distinct hotspots spanning the 3’ end of the genome and predominantly mapping to the sense strand. Therefore, it appears that the subgenomic RNA of RRV drives vpiRNA production in Ae. aegypti.

To investigate if the vpiRNAs in this study showed the U1/A10 nucleotide bias characteristic of ping-pong amplification, we used Weblogo software to determine the nucleotide frequency at each position for reads grouped by the same length, strand, and time-point post infection. There was a modest A10 bias in fat body anti-sense sequences of 28 nt at 6 dpi (Figure 2.12A), whereas the corresponding U1 bias was more readily apparent (Figure 2.12B).

We used the Small RNA Signatures tool from the Galaxy instance Mississippi.sorbonne- universite.fr to investigate 10 nt overlap probabilities in sequences in the range of 27-29 nt that mapped to the RRV genome. The software pairs sequences with their reverse complement and tally the number of pairs of sequences by the number of their overlapping nucleotides. In fat body tissue at 6 dpi, there was a high frequency of sequence pairs with a 10 nt overlap (Figure 2.13). The prominent 10 nt peaks for both the number of pairs counted and the probability of seeing pairs with a 10 nt overlap are signatures of piRNA production by ping-pong amplification.

2.4 Conclusions

In summary, we found that RRV promoted the significant differential expression of 14 miRNAs in Ae. aegypti, with most occurring in the fat body tissue at 2 dpi. Several genes related to immunity were predicted as targets for the SDE miRNAs from this study, suggesting that RRV may elicit an active defence response by Ae. aegypti. A prominent increase in 21 nt sequences was apparent at 6 and 12 days post infection, indicating a strong siRNA defence response in Ae. aegypti infected with RRV. There was a detectable vpiRNA response to RRV in Ae. aegypti fat body, and those sequences showed a distinct U1 bias, although the A10 bias in the corresponding sequences was not as distinct. The overlap probabilities of the 27-29 nt sequences from fat body piRNA-like reads at 6 dpi did show a z-score spike at 10 nt, which provided a complement to the evidence from this study for RRV- derived vpiRNA production in Ae. aegypti. This study provides insight into the small RNA responses of Ae. aegypti to RRV, and in particular we provide the first characterization of the miRNA responses in Ae. aegypti to a previously unexplored old-world alphavirus in this context.

42 Acknowledgments: We wish to thank Rhys Parry from the University of Queensland for helpful discussions in regard to data analysis, and Jody Peters from UQ for providing the virus.

Supplementary files S1-6 are available at https://www.mdpi.com/1999-4915/12/7/695/s1.

Chapter 3 Effect of miRNAs on RRV replication in Aag2 cells

Investigating the effect of differentially expressed microRNAs on Ross River virus replication in Aedes aegypti-derived Aag2 cells

3.1 Introduction

The overarching role of microRNAs (miRNAs) are for the post-transcriptional silencing of gene expression, with the principal effect of contributing to the coordinated maintenance of homeostatic balance (Gurtan and Sharp, 2013; Hussain and Asgari, 2014). Stimuli that coerce adjustments in gene expression may relate to any number of factors including responses to changes in temperature or nutrient availability, or for responses to pathogen invasion (Etebari et al., 2017). The network of genes that are differentially expressed in response to infection may involve the upregulation of genes that are directly involved in a defence response or for the suppression of circumstantially non- essential processes such as fecundity (Schwenke et al., 2016). The relationship between the systemic adaptations to viral infection and the differential abundance of miRNAs pertains to the relative expression of their target genes that may be invoked or suppressed during the process of reclaiming homeostasis (Boosani and Agrawal, 2016). Under the condition of infection, the miRNAs that would otherwise target genes that are directly involved in the defence response may be concertedly depleted and, conversely, an enrichment of miRNAs for the targeted silencing of unessential gene activity would follow. The role of miRNAs in the replication of viruses largely relates to a quashing of unessential gene activity so as to prioritise defence resources (Schwenke et al., 2016). Conversely, the trading of defence resources for the prioritisation of gene activity involved in fundamental processes, such as the digestion of blood, causes a temporary weakening of the immunity which can be exploited. In Aedes aegypti, Hussain et al. (2013b) showed that the blood-meal induced enrichment of miR-275, that targets the immunity related genes cactus and Rel1, enables the unassailed replication of dengue virus 2 (DENV2). Processes that involve innate immunity pathways are reciprocally antagonistic to fecundity and miRNAs have been shown to be key components in transitioning between these processes. In Ae. aegypti, miR-277 fine-tunes the expression of insulin- like peptides 7 (ILP7) and 8 (ILP8) that control the respective deposition and mobilisation of lipids which are important sources of energy for ovarian development. When infected, the subsequent

44 depletion of miR-277 coincides with an excessive accumulation of lipid droplets which disenables fecundity and causes an increase in the expression of FOXO (Ling et al., 2017) and the induction of Toll and IMD defence pathways which have been shown to impede DENV and Sindbis virus (SINV) replication (Barletta et al., 2016).

The proceeding study supplements the findings in Chapter 2 by extending the bioinformatic analysis to explore the differential expression of genes involved in defence responses and to parse the effect of subsequent changes in miRNA expression on Ross River virus (RRV) replication. Expression studies that show miRNA depletion or enrichment during viral infection demonstrate the correlative effect of viral insult on the system which may be an indirect result of the dynamic changes in gene expression that occurs in response to infection. The purpose of the following examination is to determine if the changes in miRNA levels that occur during viral infection correlate with increased or decreased viral replication. Such findings would lend to the hypothesis that miRNAs are involved in the coordinated response to infection playing a role in conditioning the activity of genes involved in the innate immune response. This hypothesis is strengthened by showing that the predicted target genes, of enriched miRNAs, are depleted in the same RNA-Seq file and known to be involved in innate immunity pathways. An interesting endeavour for future research would be to explore this effect as a proof of concept in cells and to be ultimately validated in mosquitoes.

3.2 Materials and Methods 3.2.1 Gene and microRNA differential expression analysis

The miRPro pipeline (Shi et al., 2015), which utilises Novoalign (Novocraft, Selangor, Malaysia) for mapping, was used to generate miRNA counts of reads mapping to the most recent miRBase repository (v21) (Griffiths-Jones et al., 2006) (Chapter 2) and SAM files of reads that aligned to the Ae. aegypti genome (AaegL5.2, VectorBase) (Giraldo-Calderón et al., 2015). Gene counts were compiled from SAM files using mirpro_feature_pro; an auxiliary program of miRPro that operates in conjunction with the subprograms HTSeq (HTSEQ 1.11) (Anders et al., 2015) and Samtools (Li et al., 2009). Differential gene expression analysis was conducted using the R program DESeq2 (v3.11) (Love et al., 2014) by default parameters using an adjusted p-value of 0.1 according to the Benjamini and Hochberg algorithm (Benjamini and Hochberg, 1995). The DESeq2 function, approximate posterior estimation for generalized linear model (apeglm) (Zhu et al., 2019); an application that discretionally shrinks the fold changes of lowly expressed genes, was used for filtering. Remaining genes with an adjusted p-value < 0.1 and fold changes > 1.5 or < -1.5 were considered as significantly differentially expressed (SDE). The SDE miRNAs that were identified in Chapter 2 were incorporated into this study. Those results were also obtained with DESeq2 using 45 significance thresholds of an adjusted p-value of 0.05, fold changes > 1.5 or < -1.5, and a normalised cross-sample mean count (baseMean) > 100.

3.2.2 Reciprocally co-expressed microRNA and target gene analysis

The putative miRNA target genes of the significantly differentially enriched (miR-9a-5p, miR- 9b-5p, miR-989-5p, miR-2940-3p and miR-2945-3p) or depleted (miR-8-5p, miR-34-3p, miR-71-3p, miR-184-3p, miR-275-3p, miR-275-5p, miR-317-3p, miR-988-3p and miR-2940-5p) miRNAs, determined in Chapter 2, were determined using the consensus from at least 3 of 5 target prediction algorithms; miRanda (Enright et al., 2003), RNAhybrid (Krüger and Rehmsmeier, 2006), TargetScan (Agarwal et al., 2018), PITA (Kertesz et al., 2007), and RNA22 (Loher and Rigoutsos, 2012). The predicted target genes of the SDE miRNAs that were found in each sample were cross referenced with the SDE genes in each respective sample. The significant miRNAs found in each sample were grouped according to their respective direction of fold change and paired with their predicted target genes. Where more than one miRNA in a group was predicted to target a single gene, the miRNA with the most prediction algorithms in consensus was selected as representative of the most probable miRNA-target-gene pair. Where two or more miRNAs from different fold change groups were predicted to target a single gene or more than one miRNA was predicted to target a single gene each by the same number of target prediction algorithms, the miRNA with a direction of fold change opposite to the direction of fold change of the predicted target gene was selected. Where more than one miRNA with the same direction of fold change was each predicted by the same number of algorithms to target a single gene with the same direction of fold change, the miRNA with the highest record of pairings with other genes was selected. The record of miRNA-target-gene pairs was then categorised according to the different fold-change combinations in each pairing. Using the fold changes of the miRNA first and the fold changes of the paired target gene second, the categories included UP-DOWN (UD), DOWN-UP (DU), UP-UP (UU), and DOWN-DOWN (DD). With the overarching role of miRNAs being for the silencing of post-transcriptional gene expression, the category of enriched miRNAs paired with depleted genes (UD) was followed most intently. Further, pairs in this category found to have Gene Ontology (GO) or Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways terms (below) related in any way to a defence response to infection were cross-referenced with transcriptome studies involving downregulated genes in the context of viral infection.

46 3.2.3 Gene ontology enrichment analysis of differentially expressed genes

The R program topGO was used for the analysis of GO terms related to the categories biological processes (BP), molecular function (MF), and cellular components (CC) (Alexa et al., 2006). The Kolmogorov-Smirnoff test statistic was incorporated using the topGO elimination algorithm to generate p-values of the enriched GO terms of SDE genes. The Kolmogorov-Smirnoff test was used for a pairwise comparison of probabilities based on the maximum distance between the p-values of the differentially expressed genes associated with or not associated with a particular GO term. The test was used to generate a probability of likelihood that the p-values of genes in association are less than (more significant) the p-values of genes not associated with a particular GO term. Furthermore, the list of genes in each sample that was predicted to be miRNA targets were analysed for GO term enrichment and compared by pairwise probability to the parent list of SDE genes to test the likelihood that the p-values of target genes associated with a particular GO term are less than the p-values of genes in the parent list that were also associated with that GO term. The GO terms specifically related in some way to an effect of the condition of infection were screened in this examination to identify miRNA-gene-pairs that are potentially involved in defence responses.

3.2.4 KEGG pathways analysis

The R program KEGGREST was used to identify the predominant KEGG pathways assigned to the SDE genes (Tenenbaum, 2019). The Wilcoxon rank-sum test was used for the pairwise comparison of probability that the p-values of genes associated with a particular KEGG pathway are less (more significant) than the p-values of genes not associated with a particular KEGG pathway. For this analysis, the gene IDs of VectorBase needed to be converted to Ensembl gene IDs which forced the exclusion of a number of viable genes from the analysis that did not have a corresponding Ensembl ID. As with the GO enrichment analysis, the partitioned list of target genes was compared with the parent list for the pathway enrichment analysis of genes putatively regulated by miRNAs, and the pathways involved in any way with a defence response were again the predominant focus.

3.2.5 miRNA mimic/inhibitor transfection and Ross River virus inoculation

Ae. aegypti (Aag2) cells were cultured in 6-well plates at 28C in a 1:1 v/v mixture of Schneider's Drosophila Medium (Life Technologies, California, USA) and Mitsuhashi and Maramorosch (Himedia, Mumbai, India) medium supplemented with 10% foetal bovine serum (FBS) and 1% penicillin-streptomycin mix. The corresponding mimics or inhibitors (GenePharma, Shanghai) of the respective depleted or enriched miRNAs, determined in Chapter 2, were transfected

47 into replicate mosquito cell cultures (N = 3) and allowed to take effect for 3 days. For the controls, a negative control mimic, negative control inhibitor and cultures with Cellfectin II only were substituted. After 3 days in incubation, cells were partitioned, and half were re-seeded into 12-well culture plates, allowed to settle for 3 hours, and then inoculated with RRV at an MOI of 0.1. After 2 hours, the cell monolayers were washed with phosphate buffered saline (PBS) and returned to incubation at 28C in a 1:1 v/v mixture of Schneider's Drosophila Medium (Life Technologies, California, USA) and Mitsuhashi and Maramorosch (Himedia, Mumbai, India) medium supplemented with 2% foetal bovine serum (FBS) for an additional 3 days. Afterwards, the supernatant briefly clarified by centrifugation at 350 g for 4 min and stored at -80C until later use for titration by Western plaque assay. The cells of both assays were washed in PBS and stored at - 80C until use.

3.2.6 RNA extraction and miRNA RT-qPCR

Total RNA was extracted using Qiazol as per the manufacturer’s protocol. The miScript II RT kit (Qiagen, Hilden, Germany) was used following the manufacturer’s protocol to synthesize cDNA from 500 ng RNA per 20 µL reaction using the 5× HiSpec Buffer and the proprietary-sequence universal primer on a Veriti thermal cycler (Applied Biosystems, California, USA). cDNA samples were diluted in 200 µL nuclease-free water and 2 µL was amplified in 10 µL duplicate reactions with miRNA-specific primers (Integrated DNA Technologies, Iowa, USA) by RT-qPCR using miScript SYBR Green PCR kit reagents (Qiagen, Hilden, Germany) on a 72-well Rotor-Gene Q thermocycler (Qiagen). The thermocycling parameters used for amplification were 95C for 15 min, followed by 40 cycles of 94C for 15 s, 60C for 30 s, and 70C for 30 s. All qPCR runs were analysed for melt curve. The mean normalized expression (MNE) of samples was calculated using the second derivative maximum and corresponding amplification efficiencies generated by the Rotor-Gene Q software (Pfaffl, 2001). The Ae. aegypti small nuclear RNA U6 was used as the calibrator. Three biological replicates were used per treatment.

3.2.7 RRV RT-qPCR

For RNA extracted from virus treated cells, cDNA was synthesized using mMuLV reverse transcriptase as per the manufacturer’s instructions (New England Biolabs). Viral sequences were amplified using the Quantifast SYBR system (Qiagen) by RT-qPCR on a Rotor-Gene Q (Qiagen, Hilden, Germany) machine with ribosomal protein subunit 17 (RPS17) and in-house-designed RRV- specific primers (Table S1). The thermocycling parameters used for amplification were 94C for 10 min, followed by 40 cycles of 94C for 15 s and 60C for 35 s. All qPCR runs were analysed for melt 48 curve. Mean normalized expression (MNE) of samples relative to the Ae. aegypti ribosomal protein subunit 17 (RPS17) was calculated using the second derivative maximum take-off and amplification efficiency values produced by the Rotor-Gene Q software.

3.3 Results and discussion 3.3.1 Correlation analysis of differentially co-expressed microRNAs and genes

After mapping cellular miRNAs, and RRV vsiRNAs and vpiRNAs, the remaining reads ( 70- 80%) mostly mapped to Ae. aegypti protein coding genes. While the sequence data generated were small RNA sequences, we thought we could use the data to capture representation of coding genes in the samples. This approach has been used previously in pan-RNA-Seq analyses (Giraldez et al., 2019; Xu et al., 2019a; Zhao et al., 2020). The R program DESeq2 was used to estimate changes in gene expression in the Ae. aegypti fat body and midgut samples at 2, 6, and 12 days post infection (dpi) with RRV. The counts of SDE miRNAs were mostly proportional to the respective counts of the SDE genes within and between the samples throughout the time course, with the highest numbers in both cases found in fat body samples at 2 dpi (Figure 3.1). In those samples, the largest difference between the proportions of significant miRNA and gene counts related to the respective directions of fold change. Curiously, ~ 1/3 of the SDE miRNAs were upregulated and ~ 1/3 of the SDE genes were downregulated, which is suggestive of a possible interrelationship between these miRNAs and genes based on their relative abundance.

In the midgut samples at 12 dpi there was 12 SDE genes found and in the fat body samples at 6 dpi there was 11 SDE genes found. There was no SDE genes found in the midgut samples at 6 dpi. The majority of SDE genes were found in the fat body samples at 2 dpi (1278) followed by the midgut samples at 2 dpi (211) and then the fat body samples at 12 dpi (31). A number of SDE genes were found to be commonly significant among different samples (Figure 3.2). In Chapter 2, the fat body samples at 2 and 12 dpi, and the midgut samples at 2 dpi, were the only samples in which SDE miRNAs were found. Of the SDE genes found in the fat body samples at 2 dpi, 42 genes were among those predicted to be targets of the 12 corresponding SDE miRNAs found, 4 genes were predicted to be targeted by the SDE miRNA (miR-9b-5p) in midgut samples at 2 dpi, and only 1 gene was predicted to be the target of miR-71-3p in the fat body samples at 12 dpi (Figure 3.2 and Figure 3.3).

3.3.2 Gene Ontology enrichment analysis

The R program topGO was used to explore the enrichment of GO terms associated with SDE genes in Ae. aegypti fat body and midgut samples at 2, 6, and 12 dpi with RRV. We prioritised our analysis of the GO terms enrichment for depleted genes that were predicted as targets of the samples upregulated miRNAs (see Figure 3.3), especially for terms with relevance to the conditional effect of

52 infection. The threshold standards of significance we set included only downregulated genes with an adjusted p-value < 0.05. We used the default settings of topGO for the enrichment analysis which imposed an additional, more stringent, cut-off of an adjusted p-value of < 0.01. Given the heightened criteria for significance and due to the limited list of genes that met our criteria and were used in the topGO pipeline, there was no GO terms for which the accompanied annotated genes were considered significant. Therefore, since we were unable to claim any enrichment of GO terms for that subset of genes, we resorted to presenting those results as an enumeration of the annotated terms for which the associated genes were significantly differentially expressed but not necessarily overrepresented according to an expectation based on random chance. It is possible that, had we given the entire list of predicted target genes inclusive of those that were not significantly depleted, the comparative difference between that larger range in supplied p-values, may have given significant results (Timmons et al., 2015). The miRNAs that were determined in Chapter 2 to be significantly upregulated included 4 miRNAs in the fat body samples at 2 dpi, miR-9b-5p in the midgut samples at 2 dpi, and miR-2940-3p in the fat body samples at 12 dpi. None of the significantly depleted fat body genes at 12 dpi were a predicted target of miR-2940-3p and only 2 depleted midgut genes were predicted as a target of miR-9b-5p. Among the 13 depleted genes predicted as targets of the upregulated miRNAs in the fat body samples at 2 dpi, miR-9b-5p was paired with 10 genes, miR-9a- 5p paired with 1, and miR-989-3p was paired with 2 genes. Although not considered enriched the top 2 most significant GO terms for each of the categories Biological Process (BP), Molecular Function (MF), and Cellular Component (CC), were indirectly related to a defence response (Figure 3.4). The most significant protein activities occurred with respects to a cellular anatomical entity (GO:0110165) in the extracellular region (GO:0005576) with the molecular functions of phosphatase (GO:0008138) and metal iron binding (GO:0046872) activities involved in the biological processes of protein dephosphorylation (GO:0006470) and oxidation-reduction (GO:0055114), respectively. In Ae. aegypti, both metal iron binding and oxidation-reduction processes have been shown to be factors associated with defence strategies against viral infection (Cheng et al., 2016).

Figure 3.4 The enrichment analysis of the Gene Ontology (GO) terms of significantly differentially expressed Aedes aegypti fat body genes at 2 dpi with Ross River virus; genes that were predicted to be targets of the significantly upregulated miRNAs in the same tissue samples. GO terms are derived from the ontology categories (A) biological processes (BP, GO:0008150), (B) molecular function (MF, GO:0003674), and (C) cellular component (CC, GO:0005575). The most significant GO terms (red boxes) are shown in association with the relating sister (below) or parent (above) nodes. 54 To explore other potential miRNA/gene interactions outside of the more commonly reported interactions explored above, we included target genes that were both putatively up or down-regulated by miRNAs that were also either up or down-regulated. Unsurprisingly, the only significant results came from the fat body samples at 2 dpi. In total, there were 79 BP terms annotated to at least one predicted target gene that was found to be differentially expressed with an adjusted p-value < 0.01. Among those BP terms 7 were found to be significantly enriched with p-values < 0.05. Given the relaxed stringency we implemented for the expanded list of supplied genes, it was surprising to see that 5/7 terms had the terminology ‘negative regulation’, with the 5th most significant term (p-value 0.016) being negative regulation of gene expression (GO:0010629). Among the 79 terms, a number of them were particularly relevant to defence responses with some specifically mentioning innate immunity toll pathways (Figure 3.5).

Biological Process

negative regulation of MyD88-independent toll-like receptor signaling pathway

regulation of MyD88-independent toll-like receptor signaling pathway

negative regulation of toll-like receptor signaling pathway

MyD88-independent toll-like receptor signaling pathway

regulation of toll-like receptor signaling pathway

toll-like receptor signaling pathway

cellular iron ion homeostasis

metal ion homeostasis

iron ion homeostasis

iron ion transport

gene silencing by RNA

gene silencing

negative regulation of gene expression

oxidation-reduction process

response to stress

0 1 2 3 4 Annotated

Figure 3.5 Enriched Gene Ontology terms of Aedes aegypti fat body genes at 2 dpi with Ross River virus and predicted as targets of differentially expressed miRNAs in the same samples. The number of annotated genes were found to be significantly differentially expressed with p-values < 0.01. The blue bars represent the topGO analysis of enrichment p- values < 0.08 and the green bars have p-values < 0.05. 55

An analysis of GO term enrichment was also conducted for the parent list of differentially expressed genes in order to gain an overall appreciation of the main biological processes occurring with respect to GO terms involved in a defence response. The same thresholds of significance criteria were implemented as above. In total, there was 547 GO terms representing at least one SDE gene of p-value < 0.01. Among the 547 GO terms 84 were determined as significantly enriched having a p- value < 0.05. Overall, there were many SDE genes annotated by GO terms related in some way to a defence response and some of those terms were also determined to be significantly enriched (Figure 3.6).

Biological Process

gene silencing by miRNA ns

Notch signaling pathway ns

negative regulation of MyD88-independent toll-like receptor signaling pathway ns

negative regulation of toll-like receptor signaling pathway ns

MyD88-independent toll-like receptor signaling pathway ns

'de novo' AMP biosynthetic process 0.008

AMP metabolic process ns

AMP biosynthetic process 0.042

defense response to other organism ns

defense response ns

humoral immune response ns

immune system process ns

immune response ns

negative regulation of Wnt signaling pathway ns

cell-cell signaling by wnt ns

'de novo' IMP biosynthetic process 0.006

IMP biosynthetic process 0.011

IMP metabolic process 0.011

0 1 2 3 4 Annotated

56 One of the most significant GO terms was ‘de novo’ IMP biosynthetic process (GO:00061898), which relates to defence responses to viral infection by the induction of apoptosis. In Ae. aegypti apoptosis is prevented by inhibitor of apoptosis (IAP) proteins that bind to and prevent the activation of the caspase Dronc which is perpetually being stimulated to action by the adaptor protein Ark. If apoptosis is to proceed, the IAP antagonist proteins Michelob_x (Mx) and inosine monophosphate (IMP) intercept IAP enabling the activation of Dronc which causes the upregulation of the effector caspases CASPS7 and CASPS8 that in turn finalise the signalling for cell death (Wang and Clem, 2011). In Ae. aegypti, Sindbis virus (O'Neill, 2013; Sanders et al., 2005), dengue virus (Ocampo et al., 2013), and Chikungunya virus (Tchankouo-Nguetcheu et al., 2012) have been shown to modulate the expression of IMP proteins with respects to apoptotic defence responses. Another significant GO term was ‘de novo’ AMP biosynthetic process (GO:0044208). Antimicrobial peptides have been researched intently and are widely known to be stimulated to action by several innate immunity pathways involving signalling via the receptors Toll, IMD, Vago, and Dome (Cheng et al., 2016; Lee et al., 2019; Sim et al., 2014).

3.3.3 KEGG pathways enrichment analysis

The R program KEGGREST was used to analyse the enrichment of KEGG pathways associated with SDE genes of an adjusted p-value (padj) < 0.1. The samples that fell within the significance threshold and that were applied to this analysis included fat body and midgut tissues at 2 and 12 dpi, and fat body tissues at 6 dpi. Additionally, a subset of genes that were predicted to be targets of the corresponding SDE miRNAs determined in Chapter 2 were also analysed via the KEGGREST pipeline. These were also held to the same significance threshold which restricted the analysis to the fat body and midgut samples at 2 dpi only. As anticipated, the greatest collective number of different KEGG pathway annotations were for fat body genes at 2 dpi (99), which included the annotations from the corresponding subset of miRNA target genes (12). There were 46 annotations ascribed to genes from midgut samples at 2 dpi which included only 1 annotation from a single miRNA target gene. The remaining annotations, which were not found to be significantly enriched, were from midgut and fat body samples at 12 dpi, and fat body samples at 6 dpi, with 10, 4, and 2 annotations, respectively. From among the annotations ascribed to fat body genes at 2 dpi, 10 had p-values less than 0.1, which included 6 with p-values less than 0.05 (Figure 3.7). Of these, the most significant KEGG pathway, MAPK signalling pathway – fly (p-value 0.006), was also attributed to a miRNA target gene in the fat body samples 2 dpi (p-value 0.07), and midgut genes at 2 (p-value 0.059) and 12 dpi (p-value 0.75).

An enrichment analysis of depleted genes that were predicted by at least 3/5 software programs to be targets of the upregulated miRNAs in the corresponding samples was not particularly revealing at the thresholds of significance applied in this study. Only 3 pathways were returned each with only a single annotated gene and all from the fat body samples at 2 dpi. These pathways, which were not considered to be significantly enriched, included Glycolysis / gluconeogenesis (p-value 0.23), phosphatidylinositol signaling system (PI3K: p-value 0.69), and starch and sucrose metabolism (p- value 0.92). To explore other potential miRNA/gene interactions beside the canonical relationship above, we investigated the KEGG pathways of predicted target genes that were either up or downregulated and for which were predicted as targets for miRNAs that were also either up or downregulated. The same thresholds of significance criteria were implemented as above and this time the results were much more implicative of pathway enrichments specifically relating to a defence response. Interestingly, even without prior knowledge of the conditional treatment of the samples, an

58 unbiased first glance at the data shows an obvious conditional response to an infection of some description, particularly in fat body samples at 2 dpi (Figure 3.8).

3.3.4 The effect of miRNA expression on RRV replication

With reference to a number of miRNAs that were determined in Chapter 2 to be significantly differentially expressed due to RRV infection in mosquitoes, the aim of this experiment was to explore the effect of changes in miRNA expression on RRV replication in mosquito cells. The rationale of this experimental procedure is based on an expectation that if the differential expression of a miRNA has an effect on RRV replication, a pairwise comparison between samples controlling for the expression of that miRNA will show a difference in the relative quantity or titre of viral copies between treated and untreated samples. For instance, if a miRNA was determined to be significantly increased in RRV infected samples compared to control samples, it was anticipated that the inoculation of RRV in cells would also cause an increase in that miRNA. If, hypothetically, the

59 increased expression of a particular miRNA causes an impediment to RRV replication, then the controlled reduction in expression of that miRNA should enable the unimpeded proliferation of the virus, resulting in a comparatively increased titre. The controlled reduction in expression of the miRNA is achieved by transfecting the cells with miRNA inhibitors. This will bring the expression of that miRNA below baseline, such that when those cells are inoculated with RRV, their capacity to utilise the miRNA in some way to impede the replication of the virus will be limited. Even under these controlled conditions it is anticipated that, as with what was seen in mosquitoes, a response to viral infection would result in the increased expression of that miRNA. Although, by comparison to untreated cells, the final expression that is reached at the end of the experimental period is expected to be lower. Given that outcome, the titre reached by the virus in untreated cells should be lower than the titre reached in cells treated with the miRNA inhibitors. However, the effect of miRNA inhibition on viral replication may be obscured by other factors. The effectiveness of inhibition may only reduce the baseline expression by a small margin (e.g. 80% of normal) and the relative impact of this reduction on viral replication may, therefore, be too subtle and undetectable. In this study, the increase in miRNA expression following mimic treatment for miR-2940-5p and miR-275-5p was significant, as was the effect of the inhibitor for miR-2940-3p, although the decreased expression of miR-989-3p was not determined to be significant, possibly due to the relatively low overall expression levels in the Aag2 cell line (Figure 3.9). This experiment was repeated twice and both attempts failed to induce inhibition of miR-9b-5p (data not shown). This is unfortunate since that miRNA was determined in Chapter 2 to have the greatest level of differential significance among all the miRNAs explored.

Other factors that could obscure the effect of miRNA inhibition on viral replication include the response of the treated cells to viral infection which may involve a rapid increase in miRNA expression returning from below baseline to levels comparable to untreated cells, such that a pairwise comparison of viral titres between samples may prove to be insignificant. In this scenario, the timing of the measurements would be important in catching an effect. To maximise the outcome in showing an effect, the inhibition of the miRNA in cells, prior to inoculation with virus, should be substantial and the time at which measurements are taken should be comparable to those that were taken in mosquitoes. Conversely, in a similar manner, if the overexpression of a miRNA happens to be pro- viral, the cells pre-treated with miRNA inhibitors, may result in a reduction in viral replication and

61 cells not treated with miRNA inhibitors, would respond to infection by overproducing the miRNA and promoting an increase in viral replication. In this study, the time course of infection in cells spanned 3 days, whereas the time course of infection explored in Chapter 2 for mosquitoes included measurements taken at 2dpi, 6dpi, and 12dpi. The majority of the differentially expressed miRNAs that were seen in mosquitoes were found in the fat body tissue at 2dpi, and only 2 miRNAs were found in the fat body at 12dpi.

Here, we chose to explore miR-989-5p, miR-2940-5p, and miR-275-3p that had significant changes in expression in the fat body tissue at 2dpi (Figure 3.1A) and miR-2940-3p which was significantly enriched in the fat body at 12dpi. It is difficult to reconcile our rationale in pursuing the effect of the enrichment of miR-2940-3p on RRV replication during a 3 day time course in cells, where its significant enrichment in mosquitoes was only seen later on at 12dpi, and although not significant, this miRNA was specifically validated by RT-qPCR to be depleted in both fat body and midgut tissues at 2dpi (see Figure 2.6). Furthermore, the significant enrichment of miR-989-5p in the fat body at 2dpi that was determined by a differential expression analysis of RNA-Seq data was actually shown to be depleted when measured by RT-qPCR. If the actual direction of fold change of this miRNA is indeed depleted, as the RT-qPCR result suggests, the effect of the inhibitor assay may explain why the levels reached were so very low. The limitations of an assessment of the effect- change of miRNA levels in Aag2 cells is restricted to shorter time intervals because Aag2 cells undergo catastrophic cytopathic effects beyond ~ 5 days of infection with RRV (unpublished pilot studies from this project), and any interesting results seen in cell culture would only be a prelude and sign to elicit further investigation in mosquitoes.

The counterpart of the hypothetical scenario eluded to earlier relates to miRNAs that were determined in Chapter 2 to be significantly reduced in mosquitoes infected with RRV compared to uninfected controls. The presupposed outcomes discussed above would, therefore, be reversed in that if the effect of the decreased expression of a miRNA due to infection is anti-viral, the increased expression of a miRNA in cells that are pre-treated with miRNA mimics may result in an increase in viral replication compared to untreated cells. Conversely, if the decreased expression of a miRNA in response to viral infection is pro-viral, its increased expression governed by mimic treatment, may result in a decrease in viral titre. In either case, beforehand and here, the ultimate scope of the analysis of this experimental procedure cannot determine whether any changes in miRNA expression are in response to or, somehow, coerced by the virus, nor whether any effect of the relative expression of the miRNA on viral replication is dependent on its level of expression, the number of viral copies present, or indeed the ratio of its expression compared to the relative degree of viraemia. In this case,

62 the levels of expressed miRNAs in response to an early, low-titre-infection may be less impactful or revealing than the potentially decapacitating effect on viral replication higher miRNA abundances may show during late-stage infection or if viral copy numbers are otherwise high. It would be interesting to pursue this question by comparing the levels of titre between different families or strains of virus at which time, based on the degree of viraemia, a particular miRNA undergoes programmed differential expression due to an arbitrary threshold of viraemia. This research endeavour may reveal why certain miRNAs seem to be more relevant for some species of virus rather than others. The observed relevance of a particular miRNA for alphaviruses but not flaviviruses, for instance, may be due to the rapidity at which alphaviruses reach high titres compared to flaviviruses. Regarding the pilot studies of the Western plaque assays that were conducted in this project, we determined that the plaques caused by RRV in Vero cells can be detected in as early as only 12 hours, whereas for the flavivirus DENV2, plaques are formed after days not hours. In this sense a particular miRNA may be relevant for alphaviruses at an early stage of infection and then only for flaviviruses later on in infection.

3.4 Conclusion

In this study, it was perhaps surprising to find that none of the miRNAs tested seemed to have a significant effect on the replication of RRV (Figure 3.9). In sections 3.1 and 3.2, the respective analysis of GO terms and KEGG pathways clearly showed the enrichment of genes that appear to be involved in a defence response to the infection with RRV, whereas the significance of the depleted genes that were predicted as targets for the significantly upregulated miRNAs was not as clear and did not seem to have priority involvement. Overall, the changes in gene expression demonstrated a profile described as a radical and distinct defence response to infection and although there was evidence that the differential expression of certain miRNAs may play a role in finetuning the expression of genes directly involved in the defence response, the effect of their relative abundance does not seem to reveal any obvious role they play in modulating the intensity of the viraemia. In hindsight and in reserving the possibility that the miRNAs do in fact play a role, it may be that the dramatic and rapid replication of RRV may have led to an overwhelming of the cell culture assay after 3 long days, where viral numbers would probably be near peak titre. It would be interesting to repeat this experiment for an infection period of only 24 hours or less, in which an effect of miRNA differential expression on RRV replication may well be seen.

63 Table S1 Primers used in this study.

Application Primer Name Sequence RT-qPCR RRV8956S F TACAAGCACGACCCATTGCCG RRV9162AS R CATAGTCCTGCCGCCTGCTGT RPS17 qF CACTCCCAGGTCCGTGGTT RPS17 qR GGACACTTCCGGCACGTACT miScript aae-miR-275-5p CGCGCTAAGCAGGAACCGAGAC aae-miR-1-3p TGGAATGTAAAGAAGTATGGAG aae-miR-100-5p AACCCGTAGATCCGAACTTGTG aae-miR-281-5p AAGAGAGCTATCCGTCGAC aae-miR-276-3p TAGGAACTTCATACCGTGCTC aae-miR-2940-3p GTCGACAGGGAGATAAATCACT aae-miR-999-3p TGTTAACTGTAAGACTGTGTCT aae-bantam-5p CCGGTTTTCATTTTCGATCTGAC aae-miR-2945-3p TGACTAGAGGCAGACTCGTTTA aae-miR-989-3p TGTGATGTGACGTAGTGGTAC RNU6 CGCAAGGATGACACGCAAATTCGTGAAGCGTTCCATATTTTT RNA-Seq 3'adapter AGATCGGAAGAGCACACGTCTGAACTCCAGTCA miRNA aae-miR-275-5p CGCGCUAAGCAGGAACCGAGAC mimics CUCGGUUCCUGCUUAGCGCGUU Negative control UUC UCC GAA CGU GUC ACG UTT ACG UGA CAC GUU CGG AGA ATT miRNA aae-miR-9b GCAUACAGCUAAAAUCACCAAAGA inhibitors aae-miR-989-3p GUACCACUACGUCACAUCACA aae-miR-2940-3p AGUGAUUUAUCUCCCUGUCGAC Negative control CAGUACUUUUGUGUAGUACAA

64 Chapter 4 General discussion

General discussion

4.1 Introduction

In Australia, Ross River virus (RRV) is the most important cause of arboviral disease (Russell, 2002a) and research into the mosquito vectors of this virus, their involvement in transmission cycles between animal hosts (Harley et al., 2001), and of studies of the epidemiological factors that are involved in RRV outbreaks (Russell, 1994), for the purpose of controlling rates of transmission and for the prevention of infection in humans, has been ongoing for decades. Unfortunately, given the ubiquitous presence of mosquito vectors, their global spread facilitated by the importations and exportations of human travel (Agarwal et al., 2017), and due to the rapidly evolving nature of the highly mutational alphavirus genome (Holland et al., 1982; Lauring and Andino, 2010), the situation of the infestation of RRV in Australia has only become bleaker. To date, insecticides have been our best means for outbreak management, but this aged technology is fraught with problems and is largely ineffective (Tomerini, 2008). More recently science has pushed for a greater understanding of the various defence responses of the mosquito vector itself, in order to better understand the biological systems involved so as to cognate ways in which to prevent transmission in the first place (Zhang et al., 2013). One avenue of mosquito research involves the innate immunity and activation of the RNA interference (RNAi) pathways by production and differential expression of small non-coding RNAs (sncRNA), which are their predominant defence response to viral assault. The elucidation of genes involved in mosquito innate immunity has been enabled by comparative sncRNA expression analysis, however, the mosquito-virus kinetics of the small RNAs involved in these pathways remains poorly characterised. This project took advantage of this dearth in research, particularly with respects to the limited studies characterising tissue-specific changes in old-world alphaviruses and of there being no studies of the sncRNA responses in mosquitoes to RRV infection.

The comparative pairwise enquiry between control and experimental conditions is a fundamental scientific approach for elucidating biological phenomena. In the field of comparative genomics, the rise of technologies enabled by the advent of genomic sequencing has spawned this space and with it, the discovery of sncRNA and of their involvement in numerous biological processes (Asgari, 2014; Blair and Olson, 2015; Hussain et al., 2016b). Studies of comparative sncRNA expression in insects have shown that RNAi inducement by viraemia predominately involves the

65 production of virus-derived short-interfering RNAs (vsiRNA). Virus-derived piRNAs (vpiRNAs) may also be produced and, as with vsiRNA, their production positively correlates with viral titre (Miesen et al., 2016). In contrast to vsiRNAs, however, there is no evidence that vpiRNA production affects viral replication and therefore vpiRNA involvement, as a measured defence-response in targeting viral genomic RNA (vgRNA) for degradation, remains an enigma. The magnitude of microRNA (miRNA) differential abundance that may occur during infection with a virus, does not positively correlate with viral titre because, unlike vsiRNAs and sometimes vpiRNAs, miRNAs are not virus-derived, although there are some studies that demonstrate this rare and poorly understood phenomenon (Zhou et al., 2015). Because there is no link between the magnitude of miRNA differential abundance and viral titre a pairwise comparison of the relative magnitudes of the differential expression of miRNAs between samples of different viral load would be non- determinative. What can be determined and has been shown is the negative correlation between the magnitudes of miRNA differential abundance and of the relative depletion of their target genes (Bartel, 2009) that may occur during infection; although there are some examples where target genes become enriched (Asgari, 2014).

Comparing miRNAs of different species has revealed the existence of miRNA families that are widely conserved (Lee et al., 2007). The intrigue of why different pathogens may induce unique differential profiles of miRNA expression within and between species has inspired research into the miRNA target genes that are involved in innate immunity pathways for defence response (Hussain et al., 2013a), the molecular factors of the host-virus environment that may explain such similarities or dichotomies (Huang et al., 2013), or for the investigation of how and why miRNAs of a particular family arose, respond to or interact with viral fragments (Hsu et al., 2007). The characterisation of the virus-induced differential expression of miRNAs and their target genes represents a fundamental level of a closed system examination. This examination can be achieved bioinformatically using RNA-Seq data, and the advancements in technology have opened up the possibility for pan-RNA- Seq analyses for both small RNA and mRNA (Xu et al., 2019b). The RNA-Seq data introduced in Chapter 2 was prepared on an Illumina platform running a small RNA protocol, however only 10- 20% of those reads were identified as being of miRNA origin, with similar numbers also assigned to siRNA and, to a lesser extent, piRNA counts. Many reads were obligatorily ignored because these had mapped to multiple places on the genome and could not be logically assigned to any unique gene with confidence. Furthermore, a fraction of vsiRNA reads would also map to concurrent mosquito- specific viral infections but these were not explored. Overall, there still remained reads left over with many that were finally mapped to genes of the Aedes aegypti genome. It is those mappings that raise the possibility for an analysis of their differential expression between samples of different conditions.

66 Once isolated and recovered by hierarchical filtering, these reads files were used in the traditional sense by being mapped against both the transcriptome and genome for a comparative gene expression analysis. A pairwise comparison between the respective magnitudes of differential expression of the miRNAs and genes detected from within the same sample were made. An analysis of the relationship between these biotypes, especially of the Gene Ontology and KEGG pathways of the putatively targeted genes of the significantly differentially expressed miRNAs revealed interesting groups of enriched terms that related to defence responses or specific innate immunity pathways. The bioinformatic approach of cross-validating results within and between studies is a powerful means to gain incrementing confidence in the accuracy of the methodology of RNA-Seq but ultimately, the soundness of any scientific endeavour, especially within the often unpredictable and sometimes unintuitive reality revealed in the fields of natural, biological systems, bioinformatic work must be coupled with experimental validations.

As with the pairwise comparison of the dynamic shift in the profile of miRNA expressions between species against a particular pathogen, there are also differences to be expected within an organism between its different tissues or even respective types of cells. The examination of biological systems by the incorporation of cell culture is often used as a proxy for the living organism and the procedure enables the tight regulation of extraneous factors that may convolute the data, however, the incumbent drawback and irony is that those extraneous factors present in a living organism would need to be included in order to gain a holistic examination of the multiple influences that determine the range of metabolic responses to all stimuli. Interestingly, studies have shown that the effect of miRNA differential expression on viral replication, under the comparable conditions of a living organism, reproduced in a cell culture system, may not reflect an interaction between miRNA expression and viral copies measured within cells but the effect of differentially expressed miRNAs may be readily apparent when observing the titre in the supernatant (Asgari, S,. personal communication). Nevertheless, as with the incrementing confidence gained by the bioinformatic implementation of multiple cross-validating results, the addition of results of cell culture and an analysis of its interrelations may bolster a hypothesis.

In this study, the exploration of the effect of miRNA differential expression on both gene expression and RRV replication was investigated. The relative quantification of gene expression was solely conducted using bioinformatic applications and those results were used to provide meaning to the cell culture experiments. Those experiments were designed to explore the effect of applying mimics or inhibitors of select miRNAs into Ae. aegypti (Aag2) cells so as to determine their involvement in modulating the capacity for the replication of RRV. We showed through GO and

67 KEGG pathway enrichment analyses that the profile changes in gene expression are concordant with an active defence response and align with what might be an expected defence response to a prolific infection with RRV. From within the same samples, differential expression of miRNAs was also calculated, and the utility of miRNA target prediction programs informed the most probable miRNA- gene pairs. The subset of genes predicted to be targets of relevant miRNAs in each sample was channelled through the R programs topGO and KEGGREST pipelines to ascertain their status of enrichment with regards to any involvement in defence response or innate immunity terminology. To that end, the results were not clear despite the parent list of genes seemingly being involved in biological processes that would usually indicate miRNA involvement, such as post-transcriptional gene silencing and specifically some genes were annotated with the GO term ‘gene silencing by miRNA’. This may indicate the involvement of miRNAs that were not considered significantly differentially expressed in Chapter 2 and subsequently not investigated or may be an effect of the combination of multiple miRNAs targeting one or more genes in that group, which was not comprehensively addressed in this project. Noted was the success of the treatment of miRNA mimics or inhibitors in cells but the impact of their respective modulation on RRV replication was not evident. The design of the assay may explain the lack of results with the possibility that the period of time in which the virus was allowed to replicate could have been too long. In future, especially given the knowledge of the rapid growth rate of RRV, an assay suspended after only 24 hours or less, may reveal an effect of miRNA abundance of RRV replication.

A similar issue was seen with respects to the timing of the sampling of infected mosquito tissues and the overlap between the respective proteins of different pathways such as the hub protein loquacious which is preferentially used for either siRNA or miRNA processing. The utility of this protein by siRNA during infection blocks access by miRNA which effectively prevents the generation of miRNA until after the siRNA response is over and the hub protein becomes available (Haac et al., 2015). Thus, each pathway must take turns to use the protein and this system ensures that only one pathway can be operational at any one time. It would stand to reason that miRNA expression would be highest during the acute phase of viral infection when the viraemia is low but as viral copy numbers increase the hub protein would become intractably less available and towards the later phases of infection miRNA involvement would decline because of their incumbently limited production. This phenomenon was shown in this study where the majority of the changes in miRNA expression occurred early during infection and very little changes were seen in the later stages of infection. Interestingly, the fat body was the prominent site of miRNA change but not the midgut. The midgut being the first point of contact for the invading viruses would be the first organ to see infection and the fat body would be infected later on following dissemination. RRV replicates very quickly and by

68 24 hours post-infection the titre is already substantial. The time of record only began at 2 dpi by which time any change in miRNAs may have already occurred and the siRNA response displaced miRNAs from the hub protein and so only miRNA changes were observed in the fat body at day 2 when dissemination had only recently appeared.

4.2 Concluding remarks

This project was a foray into the small non-coding RNA responses of Ae. aegypti to RRV with many of the profile changes showing similarity to those elicited by other related alphaviruses, although there were still changes that seem to be induced ubiquitously by multiple species of virus. The greatest similarity to other published alphavirus reports were for the piRNA responses and of the reads coverage that formed distinct clusters around the 3’ subgenomic promotor region on the RRV genome. The meaning of these clusters suggests that subgenomic sequences drive piRNA production but the overall involvement of piRNAs in defence responses remains an enigma. In conclusion, this project showed that RRV invokes an early and profound defence response in Ae. aegypti mosquitoes, and that the respective genes and pathways recruited to action are largely different between the midgut and fat body tissues. We demonstrated an RNAi defence response of Ae. aegypti to RRV which involved the differential expression of miRNA, siRNA, and piRNA sequences. Possibly the most perplexing results from this project involved the early and prominent miRNA response which became progressively weakened toward the later stages of infection. This predominately early activity of miRNA differential expression may have been a factor in not being able to demonstrate if their relative abundance affects RRV replication in cells, where perhaps an effect may have occurred earlier but was subsequently missed upon inspection after 3 days post infection.

69 References

References

Aaskov, J., Jones, A., Choi, W., Lowry, K., Stewart, E., 2011. Lineage replacement accompanying duplication and rapid fixation of an RNA element in the nsP3 gene in a species of alphavirus. Virology 410, 353-359. Aaskov, J.G., Mataika, J.U., Lawrence, G.W., Rabukawaqa, V., Tucker, M.M., Miles, J.A., Dalglish, D.A., 1981. An epidemic of Ross River virus infection in Fiji, 1979. Am J Trop Med Hyg 30, 1053-1051-1059. Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F., 2000. The genome sequence of Drosophila melanogaster. Science 287, 2185-2195. Afgan, E., Baker, D., Batut, B., Van Den Beek, M., Bouvier, D., Čech, M., Chilton, J., Clements, D., Coraor, N., Grüning, B.A., 2018. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Res 46, W537-W544. Agarwal, A., Parida, M., Dash, P.K., 2017. Impact of transmission cycles and vector competence on global expansion and emergence of arboviruses. Rev Med Virol 27, e1941. Agarwal, V., Subtelny, A.O., Thiru, P., Ulitsky, I., Bartel, D.P., 2018. Predicting microRNA targeting efficacy in Drosophila. Genome Biol 19, 152. Akbari, O.S., Antoshechkin, I., Amrhein, H., Williams, B., Diloreto, R., Sandler, J., Hay, B.A., 2013. The developmental transcriptome of the mosquito Aedes aegypti, an invasive species and major arbovirus vector. G3: Genes, Genomes, Genetics 3, 1493-1509. Akhrymuk, I., Frolov, I., Frolova, E.I., 2018b. Sindbis virus infection causes cell death by nsP2- induced transcriptional shutoff or by nsP3-dependent translational shutoff. J Virol 92, e01388- 18. Alexa, A., Rahnenführer, J., Lengauer, T., 2006. Improved scoring of functional groups from gene expression data by decorrelating GO graph structure. Bioinformatics 22, 1600-1607. Anders, S., Pyl, P.T., Huber, W., 2015. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166-169. Angleró-Rodríguez, Y.I., MacLeod, H.J., Kang, S., Carlson, J.S., Jupatanakul, N., Dimopoulos, G., 2017. Aedes aegypti molecular responses to Zika virus: modulation of infection by the toll and jak/stat immune pathways and virus host factors. Front Microbiol 8, 2050. Arensburger, P., Hice, R.H., Wright, J.A., Craig, N.L., Atkinson, P.W., 2011. The mosquito Aedes aegypti has a large genome size and high transposable element load but contains a low proportion of transposon-specific piRNAs. BMC Genomics 12, 606. Arensburger, P., Megy, K., Waterhouse, R.M., Abrudan, J., Amedeo, P., Antelo, B., Bartholomay, L., Bidwell, S., Caler, E., Camara, F., 2010. Sequencing of Culex quinquefasciatus establishes a platform for mosquito comparative genomics. Science 330, 86-88. Asad, S., Hussain, M., Hugo, L., Osei-Amo, S., Zhang, G., Watterson, D., Asgari, S., 2018. Suppression of the pelo protein by Wolbachia and its effect on dengue virus in Aedes aegypti. PLoS Neg Trop Dis 12, e0006405. Asgari, S., 2014. Role of microRNAs in arbovirus/vector interactions. Viruses 6, 3514-3534. Atasheva, S., Fish, A., Fornerod, M., Frolova, E.I., 2010. Venezuelan equine encephalitis virus capsid protein forms a tetrameric complex with CRM1 and importin α/β that obstructs nuclear pore complex function. J Virol 84, 4158-4171. Aubry, M., Teissier, A., Huart, M., Merceron, S., Vanhomwegen, J., Roche, C., Vial, A.L., Teururai, S., Sicard, S., Paulous, S., Despres, P., Manuguerra, J.C., Mallet, H.P., Musso, D., Deparis, X.,

70 Cao-Lormeau, V.M., 2017. Ross River virus seroprevalence, French Polynesia, 2014-2015. Emerg Infect Dis 23, 1751-1753. Azuolas, J., Wishart, E., Bibby, S., Ainsworth, C., 2003. Isolation of Ross River virus from mosquitoes and from horses with signs of musculoskeletal disease. Aust Vet J 81, 344-347. Babiarz, J.E., Hsu, R., Melton, C., Thomas, M., Ullian, E.M., Blelloch, R., 2011. A role for noncanonical microRNAs in the mammalian brain revealed by phenotypic differences in Dgcr8 versus Dicer1 knockouts and small RNA sequencing. RNA 17, 1489-1501. Barletta, A.B.F., Alves, L.R., Silva, M.C.L.N., Sim, S., Dimopoulos, G., Liechocki, S., Maya- Monteiro, C.M., Sorgine, M.H.F., 2016. Emerging role of lipid droplets in Aedes aegypti immune response against bacteria and Dengue virus. Sci Rep 6, 19928. Bartel, D.P., 2009. MicroRNAs: target recognition and regulatory functions. Cell 136, 215-233. Becker, T., Loch, G., Beyer, M., Zinke, I., Aschenbrenner, A.C., Carrera, P., Inhester, T., Schultze, J.L., Hoch, M., 2010. FOXO-dependent regulation of innate immune homeostasis. Nature 463, 369-373. Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 57, 289-300. Bhatt, S., Gething, P.W., Brady, O.J., Messina, J.P., Farlow, A.W., Moyes, C.L., Drake, J.M., Brownstein, J.S., Hoen, A.G., Sankoh, O., 2013. The global distribution and burden of dengue. Nature 496, 504. Bick, M.J., Carroll, J.-W.N., Gao, G., Goff, S.P., Rice, C.M., MacDonald, M.R., 2003. Expression of the zinc-finger antiviral protein inhibits alphavirus replication. J Virol 77, 11555-11562. Blair, C.D., Olson, K.E., 2015. The role of RNA interference (RNAi) in arbovirus-vector interactions. Viruses 7, 820-843. Boosani, C.S., Agrawal, D.K., 2016. Epigenetic regulation of innate immunity by microRNAs. Antibodies 5, 8. Bowers, D.F., Coleman, C.G., Brown, D.T., 2003. Sindbis virus-associated pathology in Aedes albopictus (Diptera: Culicidae). J Med Entomol 40, 698-705. Boyd, A.M., Hall, R.A., Gemmell, R.T., Kay, B.H., 2001. Experimental infection of Australian brushtail possums, Trichosurus vulpecula (Phalangeridae: Marsupialia), with Ross River and Barmah Forest viruses by use of a natural mosquito vector system. Am J Trop Med Hyg 65, 777-782. Brackney, D.E., Correa, M.A., Cozens, D.W., 2020. The impact of autophagy on arbovirus infection of mosquito cells. PLoS neglected tropical diseases 14, e0007754. Brennecke, J., Aravin, A.A., Stark, A., Dus, M., Kellis, M., Sachidanandam, R., Hannon, G.J., 2007. Discrete small RNA-generating loci as master regulators of transposon activity in Drosophila. Cell 128, 1089-1103. Bronkhorst, A.W., van Rij, R.P., 2014. The long and short of antiviral defense: small RNA-based immunity in insects. Curr Opin Virol 7, 19-28. Brown, R.S., Wan, J.J., Kielian, M., 2018. The alphavirus exit pathway: What we know and what we wish we knew. Viruses 10, 89. Bryant, B., Macdonald, W., Raikhel, A.S., 2010. microRNA miR-275 is indispensable for blood digestion and egg development in the mosquito Aedes aegypti. Proc Natl Acad Sci, USA 107, 22391-22398. Callender, D.M., 2018. Factors contributing to and strategies to combat emerging arboviruses. Glob Public Health 13, 1846-1852. Campbell, C.L., Harrison, T., Hess, A.M., Ebel, G.D., 2014. MicroRNA levels are modulated in Aedes aegypti after exposure to Dengue‐2. Insect Mol Biol 23, 132-139. Campbell, C.L., Keene, K.M., Brackney, D.E., Olson, K.E., Blair, C.D., Wilusz, J., Foy, B.D., 2008. Aedes aegypti uses RNA interference in defense against Sindbis virus infection. BMC Microbiol 8, 47. Cheng, G., Liu, Y., Wang, P., Xiao, X., 2016. Mosquito defense strategies against viral infection. Trends Parasitol 32, 177-186.

71 Chunsheng, Z., Lianhua, J., Xinglong, Y., Weijun, C., Yinian, R., Zhenzhi, L., 1997. Isolation of Ross River virus and its antibody prevalance in Hainan Province. Chinese J Vet Sci 17, 241- 243. Ciano, K.A., Saredy, J.J., Bowers, D.F., 2014. Heparan sulfate proteoglycan: an arbovirus attachment factor integral to mosquito salivary gland ducts. Viruses 6, 5182-5197. Cirimotich, C.M., Scott, J.C., Phillips, A.T., Geiss, B.J., Olson, K.E., 2009. Suppression of RNA interference increases alphavirus replication and virus-associated mortality in Aedes aegypti mosquitoes. BMC Microbiol 9, 49. Claflin, S.B., Webb, C.E., 2015. Ross River virus: many vectors and unusual hosts make for an unpredictable pathogen. PLoS Pathog 11, e1005070. Contigiani, M.S., Diaz, L.A., 2017. Togaviridae, Arthropod Borne Diseases. Springer, pp. 115-135. Crooks, G.E., Hon, G., Chandonia, J.-M., Brenner, S.E., 2004. WebLogo: a sequence logo generator. Genome Res 14, 1188-1190. Denli, A.M., Tops, B.B., Plasterk, R.H., Ketting, R.F., Hannon, G.J., 2004. Processing of primary microRNAs by the Microprocessor complex. Nature 432, 231. Derraik, J.G., 2004. A survey of the mosquito (Diptera: Culicidae) fauna of the Auckland Zoological Park. New Zealand Entomol 27, 51-55. Dietzel, W., Massion, W., Hinshaw, L., 1969. The mechanism of histamine-induced transcapillary fluid movement. Pflügers Archiv 309, 99-106. Dodt, M., Roehr, J.T., Ahmed, R., Dieterich, C., 2012. FLEXBAR—flexible barcode and adapter processing for next-generation sequencing platforms. Biology 1, 895-905. Doherty, R., Whitehead, R., Gorman, B., O'gower, A., 1963. The isolation of a third group A arbovirus in Australia, with preliminary observations on its relationship to epidemic polyarthritis. Aust J Sci 26, 183-184. Dong, S., Balaraman, V., Kantor, A.M., Lin, J., Grant, D.G., Held, N.L., Franz, A.W., 2017a. Chikungunya virus dissemination from the midgut of Aedes aegypti is associated with temporal basal lamina degradation during bloodmeal digestion. PLoS Negl Trop Dis 11, e0005976. Dong, S., Behura, S.K., Franz, A.W., 2017b. The midgut transcriptome of Aedes aegypti fed with saline or protein meals containing chikungunya virus reveals genes potentially involved in viral midgut escape. BMC Genomics 18, 382. Elton, T.S., Selemon, H., Elton, S.M., Parinandi, N.L., 2013. Regulation of the MIR155 host gene in physiological and pathological processes. Gene 532, 1-12. Enright, A.J., John, B., Gaul, U., Tuschl, T., Sander, C., Marks, D.S., 2003. MicroRNA targets in Drosophila. Genome Biol 5, R1. Etebari, K., Asgari, S., 2014. Accuracy of microRNA discovery pipelines in non-model organisms using closely related species genomes. PLoS One 9, e84747. Etebari, K., Hegde, S., Saldaña, M.A., Widen, S.G., Wood, T.G., Asgari, S., Hughes, G.L., 2017. Global transcriptome analysis of Aedes aegypti mosquitoes in response to Zika virus infection. mSphere 2, e00456-17. Etebari, K., Osei-Amo, S., Blomberg, S.P., Asgari, S., 2015. Dengue virus infection alters post- transcriptional modification of microRNAs in the mosquito vector Aedes aegypti. Sci Rep 5, 15968. Forrester, N.L., Coffey, L.L., Weaver, S.C., 2014. Arboviral bottlenecks and challenges to maintaining diversity and fitness during mosquito transmission. Viruses 6, 3991-4004. Fragkoudis, R., Attarzadeh-Yazdi, G., Nash, A.A., Fazakerley, J.K., Kohl, A., 2009. Advances in dissecting mosquito innate immune responses to arbovirus infection. J Gen Virol 90, 2061- 2072. Fragkoudis, R., Chi, Y., Siu, R., Barry, G., Attarzadeh‐Yazdi, G., Merits, A., Nash, A., Fazakerley, J., Kohl, A., 2008. Semliki Forest virus strongly reduces mosquito host defence signaling. Insect Mol Biol 17, 647-656. Franz, A.W., Kantor, A.M., Passarelli, A.L., Clem, R.J., 2015. Tissue barriers to arbovirus infection in mosquitoes. Viruses 7, 3741-3767.

72 Franz, A.W., Sanchez-Vargas, I., Adelman, Z.N., Blair, C.D., Beaty, B.J., James, A.A., Olson, K.E., 2006. Engineering RNA interference-based resistance to dengue virus type 2 in genetically modified Aedes aegypti. Proc Natl Acad Sci, USA 103, 4198-4203. Fros, J.J., Liu, W.J., Prow, N.A., Geertsema, C., Ligtenberg, M., Vanlandingham, D.L., Schnettler, E., Vlak, J.M., Suhrbier, A., Khromykh, A.A., 2010. Chikungunya virus nonstructural protein 2 inhibits type I/II interferon-stimulated JAK-STAT signaling. J Virol 84, 10877-10887. Fros, J.J., Pijlman, G.P., 2016. Alphavirus infection: host cell shut-off and inhibition of antiviral responses. Viruses 8, 166. Fu, X., Liu, P., Dimopoulos, G., Zhu, J., 2020. Dynamic miRNA-mRNA interactions coordinate gene expression in adult Anopheles gambiae. PLoS Genetics 16, e1008765. Garbuzov, A., Tatar, M., 2010. Hormonal regulation of Drosophila microRNA let-7 and miR-125 that target innate immunity. Fly 4, 306-311. Garmashova, N., Gorchakov, R., Volkova, E., Paessler, S., Frolova, E., Frolov, I., 2007. The Old World and New World alphaviruses use different virus-specific proteins for induction of transcriptional shutoff. J Virol 81, 2472-2484. Garoff, H., Simons, K., 1974. Location of the spike glycoproteins in the Semliki Forest virus membrane. Proc Natl Acad Sci, USA 71, 3988-3992. Gel, B., Serra, E., 2017. karyoploteR: an R/Bioconductor package to plot customizable genomes displaying arbitrary data. Bioinformatics 33, 3088-3090. Ghildiyal, M., Zamore, P.D., 2009. Small silencing RNAs: an expanding universe. Nat Rev Genet 10, 94-108. Giraldez, M.D., Spengler, R.M., Etheridge, A., Goicochea, A.J., Tuck, M., Choi, S.W., Galas, D.J., Tewari, M., 2019. Phospho‐RNA‐seq: a modified small RNA‐seq method that reveals circulating mRNA and lncRNA fragments as potential biomarkers in human plasma. EMBO J 38, e101695. Giraldo-Calderón, G.I., Emrich, S.J., MacCallum, R.M., Maslen, G., Dialynas, E., Topalis, P., Ho, N., Gesing, S., Consortium, V., Madey, G., 2015. VectorBase: an updated bioinformatics resource for invertebrate vectors and other organisms related with human diseases. Nucleic Acids Res 43, D707-D713. Götz, S., García-Gómez, J.M., Terol, J., Williams, T.D., Nagaraj, S.H., Nueda, M.J., Robles, M., Talón, M., Dopazo, J., Conesa, A., 2008. High-throughput functional annotation and data mining with the Blast2GO suite. Nucleic Acids Res 36, 3420-3435. Griffiths-Jones, S., Grocock, R.J., Van Dongen, S., Bateman, A., Enright, A.J., 2006. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res 34, D140-D144. Grimley, P.M., Levin, J.G., Berezesky, I.K., Friedman, R.M., 1972. Specific membranous structures associated with the replication of group A arboviruses. J Virol 10, 492-503. Gubler, D.J., 1981. Transmission of Ross River virus by Aedes polynesiensis and Aedes aegypti. Am J Trop Med Hyg 30, 1303-1306. Gulia-Nuss, M., Robertson, A.E., Brown, M.R., Strand, M.R., 2011. Insulin-like peptides and the target of rapamycin pathway coordinately regulate blood digestion and egg maturation in the mosquito Aedes aegypti. PLoS One 6, e20401. Gulyaeva, L.F., Kushlinskiy, N.E., 2016. Regulatory mechanisms of microRNA expression. J Transl Med 14, 143. Gummow, B., Tan, R., Joice, R., Burgess, G., Picard, J., 2018. Seroprevalence and associated risk factors of mosquito‐borne alphaviruses in horses in northern Queensland. Aust Vet J 96, 243- 251. Gunawardane, L.S., Saito, K., Nishida, K.M., Miyoshi, K., Kawamura, Y., Nagami, T., Siomi, H., Siomi, M.C., 2007. A slicer-mediated mechanism for repeat-associated siRNA 5'end formation in Drosophila. Science 315, 1587-1590. Gurtan, A.M., Sharp, P.A., 2013. The role of miRNAs in regulating gene expression networks. J Mol Biol 425, 3582-3600.

73 Haac, M.E., Anderson, M.A., Eggleston, H., Myles, K.M., Adelman, Z.N., 2015. The hub protein loquacious connects the microRNA and short interfering RNA pathways in mosquitoes. Nucleic Acids Res 43, 3688-3700. Harley, D., Sleigh, A., Ritchie, S., 2001. Ross River virus transmission, infection, and disease: a cross-disciplinary review. Clin Microbiol Rev 14, 909-932. Harrison, D.A., McCoon, P.E., Binari, R., Gilman, M., Perrimon, N., 1998. Drosophila unpaired encodes a secreted protein that activates the JAK signaling pathway. Genes Dev 12, 3252-3263. Hedges, L.M., Brownlie, J.C., O'Neill, S.L., Johnson, K.N., 2008. Wolbachia and virus protection in insects. Science 322, 702. Hennessey, M., Fischer, M., Staples, J.E., 2016. Zika virus spreads to new areas—region of the Americas, May 2015–January 2016. MMWR Morb Mortal Wkly Rep 65, 55-58. Holder, P., Browne, G., Bullians, M., 1999. The mosquitoes of New Zealand and their animal disease significance. Surveillance 26, 12-15. Holland, J., Spindler, K., Horodyski, F., Grabau, E., Nichol, S., VandePol, S., 1982. Rapid evolution of RNA genomes. Science 215, 1577-1585. Holt, R.A., Subramanian, G.M., Halpern, A., Sutton, G.G., Charlab, R., Nusskern, D.R., Wincker, P., Clark, A.G., Ribeiro, J.C., Wides, R., 2002. The genome sequence of the malaria mosquito Anopheles gambiae. Science 298, 129-149. Hsu, P.W.-C., Lin, L.-Z., Hsu, S.-D., Hsu, J.B.-K., Huang, H.-D., 2007. ViTa: prediction of host microRNAs targets on viruses. Nucleic Acids Res 35, D381-D385. Huang, Z., Kingsolver, M.B., Avadhanula, V., Hardy, R.W., 2013. An antiviral role for antimicrobial peptides during the arthropod response to alphavirus replication. J Virol 87, 4272-4280. Huber, W., Carey, V.J., Gentleman, R., Anders, S., Carlson, M., Carvalho, B.S., Bravo, H.C., Davis, S., Gatto, L., Girke, T., 2015. Orchestrating high-throughput genomic analysis with Bioconductor. Nature Meth 12, 115. Hussain, M., Asgari, S., 2014. A glance on the role of miRNAs in insect life, Insect Molecular Biology and Ecology. CRC Press, pp. 130-168. Hussain, M., Etebari, K., Asgari, S., 2016a. Functions of small RNAs in mosquitoes, Adv Insect Physiol. Elsevier, pp. 189-222. Hussain, M., Etebari, K., Asgari, S., 2016b. Functions of small RNAs in mosquitoes, Advances in Insect Physiology. Elsevier, pp. 189-222. Hussain, M., Walker, T., O'Neill, S.L., Asgari, S., 2013a. Blood meal induced microRNA regulates development and immune associated genes in the Dengue mosquito vector, Aedes aegypti. Insect Biochem Mol Biol 43, 146-152. Hussain, M., Walker, T., O'Neill, S.L., Asgari, S., 2013b. Blood meal induced microRNA regulates development and immune associated genes in the Dengue mosquito vector, Aedes aegypti. Insect Biochem Mol Biol 43, 146-152. Jones, A., Lowry, K., Aaskov, J., Holmes, E.C., Kitchen, A., 2010. Molecular evolutionary dynamics of Ross River virus and implications for vaccine efficacy. J Gen Virol 91, 182-188. Joosten, J., Miesen, P., Taşköprü, E., Pennings, B., Jansen, P.W., Huynen, M.A., Vermeulen, M., Van Rij, R.P., 2019. The Tudor protein Veneno assembles the ping-pong amplification complex that produces viral piRNAs in Aedes mosquitoes. Nucleic Acids Res 47, 2546-2559. Jupatanakul, N., Sim, S., Angleró-Rodríguez, Y.I., Souza-Neto, J., Das, S., Poti, K.E., Rossi, S.L., Bergren, N., Vasilakis, N., Dimopoulos, G., 2017. Engineered Aedes aegypti JAK/STAT pathway-mediated immunity to dengue virus. PLoS Neg Trop Dis 11, e0005187. Kay, B., 1982. Three modes of transmission of Ross River virus by Aedes vigilax (Skuse). Aust J Exp Biol Med Sci 60, 339-344. Kay, B.H., Boyd, A.M., Ryan, P.A., Hall, R.A., 2007. Mosquito feeding patterns and natural infection of vertebrates with Ross River and Barmah Forest viruses in Brisbane, Australia. Am J Trop Med Hyg 76, 417-423. Kelly, E.M., Moon, D.C., Bowers, D.F., 2012. Apoptosis in mosquito salivary glands: Sindbis virus- associated and tissue homeostasis. J Gen Virol 93, 2419-2424.

74 Kelly‐Hope, L.A., Kay, B.H., Purdie, D.M., Williams, G.M., 2002. Infectious disease: the risk of Ross River and Barmah Forest virus disease in Queensland: implications for New Zealand. Aust New Zealand J Public Health 26, 69-77. Kertesz, M., Iovino, N., Unnerstall, U., Gaul, U., Segal, E., 2007. The role of site accessibility in microRNA target recognition. Nat Genet 39, 1278-1284. Khoo, C.C., Piper, J., Sanchez-Vargas, I., Olson, K.E., Franz, A.W., 2010. The RNA interference pathway affects midgut infection-and escape barriers for Sindbis virus in Aedes aegypti. BMC Microbiol 10, 130. Kim, C.-H., Muturi, E.J., 2013. Effect of larval density and Sindbis virus infection on immune responses in Aedes aegypti. J Insect Physiol 59, 604-610. Koolhof, I., Carver, S., 2017. Epidemic host community contribution to mosquito-borne disease transmission: Ross River virus. Epidemiol Infect 145, 656-666. Kraemer, M.U., Sinka, M.E., Duda, K.A., Mylne, A.Q., Shearer, F.M., Barker, C.M., Moore, C.G., Carvalho, R.G., Coelho, G.E., Van Bortel, W., 2015. The global distribution of the arbovirus vectors Aedes aegypti and Ae. albopictus. eLife 4, e08347. Krüger, J., Rehmsmeier, M., 2006. RNAhybrid: microRNA target prediction easy, fast and flexible. Nucleic Acids Res 34, W451-W454. Kuss-Duerkop, S.K., Wang, J., Mena, I., White, K., Metreveli, G., Sakthivel, R., Mata, M.A., Muñoz- Moreno, R., Chen, X., Krammer, F., 2017. Influenza virus differentially activates mTORC1 and mTORC2 signaling to maximize late stage replication. PLoS Pathog 13, e1006635. Lau, C., Aubry, M., Musso, D., Teissier, A., Paulous, S., Desprès, P., De-Lamballerie, X., Pastorino, B., Cao-Lormeau, V.-M., Weinstein, P., 2017. New evidence for endemic circulation of Ross River virus in the Pacific Islands and the potential for emergence. Int J Infect Dis 57, 73-76. Lauring, A.S., Andino, R., 2010. Quasispecies theory and the behavior of RNA viruses. PLoS Pathog 6, e1001005. Lee, C.-T., Risom, T., Strauss, W.M., 2007. Evolutionary conservation of microRNA regulatory circuits: an examination of microRNA gene complexity and conserved microRNA-target interactions through metazoan phylogeny. DNA and cell biology 26, 209-218. Lee, W.-S., Webster, J.A., Madzokere, E.T., Stephenson, E.B., Herrero, L.J., 2019. Mosquito antiviral defense mechanisms: a delicate balance between innate immunity and persistent viral infection. Parasit Vectors 12, 165. Lefkowitz, E.J., Dempsey, D.M., Hendrickson, R.C., Orton, R.J., Siddell, S.G., Smith, D.B., 2018. Virus taxonomy: the database of the International Committee on Taxonomy of Viruses (ICTV). Nucleic Acids Res 46, D708-D717. Lewis, B.P., Burge, C.B., Bartel, D.P., 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120, 15-20. Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R., 2009. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078-2079. Li, L., Jose, J., Xiang, Y., Kuhn, R.J., Rossmann, M.G., 2010. Structural changes of envelope proteins during alphavirus fusion. Nature 468, 705. Li, Y., Li, S., Li, R., Xu, J., Jin, P., Chen, L., Ma, F., 2017. Genome-wide miRNA screening reveals miR-310 family members negatively regulate the immune response in Drosophila melanogaster via co-targeting Drosomycin. Dev Comp Immunol 68, 34-45. Liang, Q., Luo, Z., Zeng, J., Chen, W., Foo, S.-S., Lee, S.-A., Ge, J., Wang, S., Goldman, S.A., Zlokovic, B.V., 2016. Zika virus NS4A and NS4B proteins deregulate Akt-mTOR signaling in human fetal neural stem cells to inhibit neurogenesis and induce autophagy. Cell Stem Cell 19, 663-671. Lindsay, M., Johansen, C., Broom, A., D’Ercole, M., Wright, A., Condon, R., Smith, D., Mackenzie, J., 1992. The epidemiology of outbreaks of Ross River virus infection in Western Australia in 1991–1992. Arbovirus Res Aust 6, 72-76.

75 Ling, L., Kokoza, V.A., Zhang, C., Aksoy, E., Raikhel, A.S., 2017. MicroRNA-277 targets insulin- like peptides 7 and 8 to control lipid metabolism and reproduction in Aedes aegypti mosquitoes. Proc Natl Acad Sci, USA 114, E8017-E8024. Liu, S., Lucas, K.J., Roy, S., Ha, J., Raikhel, A.S., 2014. Mosquito-specific microRNA-1174 targets serine hydroxymethyltransferase to control key functions in the gut. Proc Natl Acad Sci, USA 111, 14460-14465. Liu, W., Kizu, J.R., Le Grand, L.R., Moller, C.G., Carthew, T.L., Mitchell, I.R., Gubala, A.J., Aaskov, J.G., 2019. Localized outbreaks of epidemic polyarthritis among military personnel caused by different sublineages of Ross River virus, northeastern Australia, 2016–2017. Emerg Infect Dis 25, 1793. Liu, Y., Zhou, Y., Wu, J., Zheng, P., Li, Y., Zheng, X., Puthiyakunnon, S., Tu, Z., Chen, X.-G., 2015. The expression profile of Aedes albopictus miRNAs is altered by dengue virus serotype-2 infection. Cell Biosci 5, 16. Loher, P., Rigoutsos, I., 2012. Interactive exploration of RNA22 microRNA target predictions. Bioinformatics 28, 3322-3323. Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550. Lucas, K.J., Roy, S., Ha, J., Gervaise, A.L., Kokoza, V.A., Raikhel, A.S., 2015. MicroRNA-8 targets the Wingless signaling pathway in the female mosquito fat body to regulate reproductive processes. Proc Natl Acad Sci, USA 112, 1440-1445. Luplertlop, N., Surasombatpattana, P., Patramool, S., Dumas, E., Wasinpiyamongkol, L., Saune, L., Hamel, R., Bernard, E., Sereno, D., Thomas, F., 2011. Induction of a peptide with activity against a broad spectrum of pathogens in the Aedes aegypti salivary gland, following infection with dengue virus. PLoS Pathog 7, e1001252. Maharaj, P.D., Widen, S.G., Huang, J., Wood, T.G., Thangamani, S., 2015. Discovery of mosquito saliva microRNAs during CHIKV infection. PLoS Neg Trop Dis 9, e0003386. Marovich, M., Grouard-Vogel, G., Louder, M., Eller, M., Sun, W., Wu, S., Putvatana, R., Murphy, G., Tassaneetrithep, B., Burgess, T., 2004. Human dendritic cells as targets of dengue virus infection. J Invest Dermatol Symp Proc 6, 219-224. Mayer, S.V., Tesh, R.B., Vasilakis, N., 2017. The emergence of arthropod-borne viral diseases: A global prospective on dengue, chikungunya and zika fevers. Acta Trop 166, 155-163. Mayoral, J.G., Etebari, K., Hussain, M., Khromykh, A.A., Asgari, S., 2014. Wolbachia infection modifies the profile, shuttling and structure of microRNAs in a mosquito cell line. PLoS One 9, e96107. Mazzon, M., Castro, C., Thaa, B., Liu, L., Mutso, M., Liu, X., Mahalingam, S., Griffin, J.L., Marsh, M., McInerney, G.M., 2018. Alphavirus-induced hyperactivation of PI3K/AKT directs pro- viral metabolic changes. PLoS Pathog 14, e1006835. McFarlane, M., Arias-Goeta, C., Martin, E., O'Hara, Z., Lulla, A., Mousson, L., Rainey, S.M., Misbah, S., Schnettler, E., Donald, C.L., 2014. Characterization of Aedes aegypti innate- immune pathways that limit Chikungunya virus replication. PLoS Neg Trop Dis 8, e2994. Mead, E.A., Tu, Z., 2008. Cloning, characterization, and expression of microRNAs from the Asian malaria mosquito, Anopheles stephensi. BMC Genomics 9, 244. Mendes, A., Kuhn, R.J., 2018. Alphavirus nucleocapsid packaging and assembly. Viruses 10, 138. Miesen, P., Girardi, E., van Rij, R.P., 2015. Distinct sets of PIWI proteins produce arbovirus and transposon-derived piRNAs in Aedes aegypti mosquito cells. Nucleic Acids Res 43, 6545-6556. Miesen, P., Joosten, J., van Rij, R.P., 2016. PIWIs go viral: arbovirus-derived piRNAs in vector mosquitoes. PLoS Pathog 12, e1006017. Moazed, D., 2009. Small RNAs in transcriptional gene silencing and genome defence. Nature 457, 413. Montgomery, S.A., Berglund, P., Beard, C.W., Johnston, R.E., 2006. Ribosomal protein S6 associates with alphavirus nonstructural protein 2 and mediates expression from alphavirus messages. J Virol 80, 7729-7739.

76 Morazzani, E.M., Wiley, M.R., Murreddu, M.G., Adelman, Z.N., Myles, K.M., 2012. Production of virus-derived ping-pong-dependent piRNA-like small RNAs in the mosquito soma. PLoS Pathog 8, e1002470. Murota, Y., Ishizu, H., Nakagawa, S., Iwasaki, Y.W., Shibata, S., Kamatani, M.K., Saito, K., Okano, H., Siomi, H., Siomi, M.C., 2014. Yb integrates piRNA intermediates and processing factors into perinuclear bodies to enhance piRISC assembly. Cell Rep 8, 103-113. Musso, D., Gubler, D.J., 2015. Zika virus: following the path of dengue and chikungunya? The Lancet 386, 243-244. Myles, K.M., Wiley, M.R., Morazzani, E.M., Adelman, Z.N., 2008. Alphavirus-derived small RNAs modulate pathogenesis in disease vector mosquitoes. Proc Natl Acad Sci, USA 105, 19938- 19943. Nakamoto, M., Moy, R.H., Xu, J., Bambina, S., Yasunaga, A., Shelly, S.S., Gold, B., Cherry, S., 2012. Virus recognition by Toll-7 activates antiviral autophagy in Drosophila. Immunity 36, 658-667. Napoli, C., Lemieux, C., Jorgensen, R., 1990. Introduction of a chimeric chalcone synthase gene into petunia results in reversible co-suppression of homologous genes in trans. Plant Cell 2, 279- 289. Nasci, R.S., Mitchell, C.J., 1994. Larval diet, adult size, and susceptibility of Aedes aegypti (Diptera: Culicidae) to infection with Ross River virus. J Med Entomol 31, 123-126. Nene, V., Wortman, J.R., Lawson, D., Haas, B., Kodira, C., Tu, Z.J., Loftus, B., Xi, Z., Megy, K., Grabherr, M., 2007. Genome sequence of Aedes aegypti, a major arbovirus vector. Science. Nishimasu, H., Ishizu, H., Saito, K., Fukuhara, S., Kamatani, M.K., Bonnefond, L., Matsumoto, N., Nishizawa, T., Nakanaga, K., Aoki, J., 2012. Structure and function of Zucchini endoribonuclease in piRNA biogenesis. Nature 491, 284. NNDSS, 2018. National Notifiable Diseases Surveillance fortnightly summary notes - 4 August to 17 August 2018. https://www1.health.gov.au/internet/main/publishing.nsf/Content/cdnareport- fn16-18.htm (accessed on 18.19.2020). O'Neill, K.L., 2013. The role of apoptotic factors in Sindbis virus infection and replication in the mosquito vector Aedes aegypti. Kansas State University. Ocampo, C.B., Caicedo, P.A., Jaramillo, G., Bedoya, R.U., Baron, O., Serrato, I.M., Cooper, D.M., Lowenberger, C., 2013. Differential expression of apoptosis related genes in selected strains of Aedes aegypti with different susceptibilities to dengue virus. PLoS One 8, e61187. Ohtsuka, E., Kawai, S., Ichikawa, T., Nojima, H., Kitagawa, K., Shirai, Y., Kamimura, K., Kuraishi, Y., 2001. Roles of mast cells and histamine in mosquito bite-induced allergic itch-associated responses in mice. Jpn J Pharmacol 86, 97-105. Ojani, R., Fu, X., Ahmed, T., Liu, P., Zhu, J., 2018. Krüppel homologue 1 acts as a repressor and an activator in the transcriptional response to juvenile hormone in adult mosquitoes. Insect Mol Biol 27, 268-278. Okamura, K., Hagen, J.W., Duan, H., Tyler, D.M., Lai, E.C., 2007. The mirtron pathway generates microRNA-class regulatory RNAs in Drosophila. Cell 130, 89-100. Old, J.M., Deane, E.M., 2005. Antibodies to the Ross River virus in captive marsupials in urban areas of eastern New South Wales, Australia. J Wildlife Dis 41, 611-614. Oliveira, A.C., Bovolenta, L.A., Nachtigall, P.G., Herkenhoff, M.E., Lemke, N., Pinhal, D., 2017. Combining results from distinct microRNA target prediction tools enhances the performance of analyses. Front Genet 8, 59. Page-McCaw, A., 2008. Remodeling the model organism: matrix metalloproteinase functions in invertebrates. Semin Cell Dev Biol 19, 14-23. Pakpour, N., Riehle, M.A., Luckhart, S., 2014. Effects of ingested vertebrate-derived factors on insect immune responses. Curr Opin Insect Sci 3, 1-5. Paredes, A.M., Brown, D.T., Rothnagel, R., Chiu, W., Schoepp, R.J., Johnston, R.E., Prasad, B., 1993. Three-dimensional structure of a membrane-containing virus. Proc Natl Acad Sci, USA 90, 9095-9099.

77 Paredes, A.M., Simon, M.N., Brown, D.T., 1992. The mass of the Sindbis virus nucleocapsid suggests it has T= 4 icosahedral symmetry. Virology 187, 329-332. Parikh, G.R., Oliver, J.D., Bartholomay, L.C., 2009. A haemocyte tropism for an arbovirus. J Gen Virol 90, 292-296. Patel, R.K., Hardy, R.W., 2012. Role for the phosphatidylinositol 3-kinase-Akt-TOR pathway during sindbis virus replication in arthropods. J Virol 86, 3595-3604. Pearson, W.R., Wood, T., Zhang, Z., Miller, W., 1997. Comparison of DNA sequences with protein sequences. Genomics 46, 24-36. Perrone, J.B., Spielman, A., 1988. Time and site of assembly of the peritrophic membrane of the mosquito Aedes aegypti. Cell Tissue Res 252, 473-478. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-time RT–PCR. Nucleic Acids Res 29, e45. Pingen, M., Bryden, S.R., Pondeville, E., Schnettler, E., Kohl, A., Merits, A., Fazakerley, J.K., Graham, G.J., McKimmie, C.S., 2016. Host inflammatory response to mosquito bites enhances the severity of arbovirus infection. Immunity 44, 1455-1469. Platt, K.B., Linthicum, K.J., Myint, K.S., Innis, B.L., Lerdthusnee, K., Vaughn, D.W., 1997. Impact of dengue virus infection on feeding behavior of Aedes aegypti. Am J Trop Med Hygiene 57, 119-125. Powell, J.R., 2018. Mosquito-borne human viral diseases: Why Aedes aegypti? Am J Trop Med Hyg 98, 1563-1565. Quinlan, A.R., Hall, I.M., 2010. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841-842. Raquin, V., Merkling, S.H., Gausson, V., Moltini-Conclois, I., Frangeul, L., Varet, H., Dillies, M.- A., Saleh, M.-C., Lambrechts, L., 2017. Individual co-variation between viral RNA load and gene expression reveals novel host factors during early dengue virus infection of the Aedes aegypti midgut. PLoS Negl Trop Dis 11, e0006152. Ratnayake, J.T.B., 2006. The valuation of social and economic costs of mosquito-transmitted Ross River virus. Griffith University. Rezza, G., 2014. Dengue and chikungunya: long-distance spread and outbreaks in naïve areas. Pathog Glob Health 108, 349-355. Ritchie, S.A., Fanning, I.D., Phillips, D.A., Standfast, H.A., Mcginn, D., Kay, B.H., 1997. Ross River virus in mosquitoes (Diptera: Culicidae) during the 1994 epidemic around Brisbane, Australia. J Med Entomol 34, 156-159. Romoser, W., Turell, M., Lerdthusnee, K., Neira, M., Dohm, D., Ludwig, G., Wasieloski, L., 2005. Pathogenesis of Rift Valley fever virus in mosquitoes—tracheal conduits & the basal lamina as an extra-cellular barrier, Infectious diseases from nature: mechanisms of viral emergence and persistence. Springer, pp. 89-100. Romoser, W.S., Wasieloski Jr, L.P., Pushko, P., Kondig, J.P., Lerdthusnee, K., Neira, M., Ludwig, G.V., 2004. Evidence for arbovirus dissemination conduits from the mosquito (Diptera: Culicidae) midgut. J Med Entomol 41, 467-475. Rose, P.P., Hanna, S.L., Spiridigliozzi, A., Wannissorn, N., Beiting, D.P., Ross, S.R., Hardy, R.W., Bambina, S.A., Heise, M.T., Cherry, S., 2011. Natural resistance-associated macrophage protein is a cellular receptor for sindbis virus in both insect and mammalian hosts. Cell Host Microbe 10, 97-104. Russell, R., 1994. Ross River virus: disease trends and vector ecology in Australia. Bull Soc Vector Ecol 19, 73-81. Russell, R.C., 2002a. Ross River virus: ecology and distribution. Ann Rev Entomol 47, 1-31. Russell, R.C., 2002b. Ross River virus: ecology and distribution. Annu Rev Entomol 47, 1-31. Ryan, P.A., Martin, L., Mackenzie, J.S., Kay, B.H., 1997. Investigation of gray-headed flying foxes (Pteropus poliocephalus)(Megachiroptera: Pteropodidae) and mosquitoes in the ecology of Ross River virus in Australia. Am J Trop Med Hygiene 57, 476-482.

78 Sadler, A.J., Williams, B.R., 2008. Interferon-inducible antiviral effectors. Nature Rev Microbiol 8, 559-568. Saldaña, M.A., Etebari, K., Hart, C.E., Widen, S.G., Wood, T.G., Thangamani, S., Asgari, S., Hughes, G.L., 2017. Zika virus alters the microRNA expression profile and elicits an RNAi response in Aedes aegypti mosquitoes. PLoS Neg Trop Dis 11, e0005760. Sanders, H.R., Foy, B.D., Evans, A.M., Ross, L.S., Beaty, B.J., Olson, K.E., Gill, S.S., 2005. Sindbis virus induces transport processes and alters expression of innate immunity pathway genes in the midgut of the disease vector, Aedes aegypti. Insect Biochem Mol Biol 35, 1293-1307. Scapini, P., Lapinet‐Vera, J.A., Gasperini, S., Calzetti, F., Bazzoni, F., Cassatella, M.A., 2000. The neutrophil as a cellular source of chemokines. Immunol Rev 177, 195-203. Schwenke, R.A., Lazzaro, B.P., Wolfner, M.F., 2016. Reproduction–immunity trade-offs in insects. Annu Rev Entomol 61, 239-256. Shelly, S., Lukinova, N., Bambina, S., Berman, A., Cherry, S., 2009. Autophagy is an essential component of Drosophila immunity against vesicular stomatitis virus. Immunity 30, 588-598. Shi, J., Dong, M., Li, L., Liu, L., Luz-Madrigal, A., Tsonis, P.A., Del Rio-Tsonis, K., Liang, C., 2015. mirPRo–a novel standalone program for differential expression and variation analysis of miRNAs. Sci Rep 5, 14617. Shin, S.W., Kokoza, V., Ahmed, A., Raikhel, A.S., 2002. Characterization of three alternatively spliced isoforms of the Rel/NF-κB transcription factor Relish from the mosquito Aedes aegypti. Proc Natl Acad Sci, USA 99, 9978-9983. Shin, S.W., Kokoza, V., Bian, G., Cheon, H.-M., Kim, Y.J., Raikhel, A.S., 2005. REL1, a homologue of Drosophila dorsal, regulates toll antifungal immune pathway in the female mosquito Aedes aegypti. J Biol Chem 280, 16499-16507. Shrinet, J., Jain, S., Jain, J., Bhatnagar, R.K., Sunil, S., 2014. Next generation sequencing reveals regulation of distinct Aedes microRNAs during chikungunya virus development. PLoS Neg Trop Dis 8, e2616. Sim, C., Hong, Y., Vanlandingham, D., Harker, B., Christophides, G., Kafatos, F., Higgs, S., Collins, F., 2005. Modulation of Anopheles gambiae gene expression in response to o’nyong‐nyong virus infection. Insect Mol Biol 14, 475-481. Sim, C., Hong, Y.S., Tsetsarkin, K.A., Vanlandingham, D.L., Higgs, S., Collins, F.H., 2007. Anopheles gambiae heat shock protein cognate 70B impedes O’nyong-nyong virus replication. BMC Genomics 8, 231. Sim, S., Jupatanakul, N., Dimopoulos, G., 2014. Mosquito immunity against arboviruses. Viruses 6, 4479-4504. Siu, R.W., Fragkoudis, R., Simmonds, P., Donald, C.L., Chase-Topping, M.E., Barry, G., Attarzadeh- Yazdi, G., Rodriguez-Andres, J., Nash, A.A., Merits, A., 2011. Antiviral RNA interference responses induced by Semliki Forest virus infection of mosquito cells: characterization, origin, and frequency-dependent functions of virus-derived small interfering RNAs. J Virol 85, 2907- 2917. Skalsky, R.L., Vanlandingham, D.L., Scholle, F., Higgs, S., Cullen, B.R., 2010. Identification of microRNAs expressed in two mosquito vectors, Aedes albopictus and Culex quinquefasciatus. BMC Genomics 11, 119. Sluss, H.K., Han, Z., Barrett, T., Davis, R.J., Ip, Y.T., 1996. A JNK signal transduction pathway that mediates morphogenesis and an immune response in Drosophila. Genes Dev 10, 2745-2758. Spellberg, M.J., Marr, M.T., 2015. FOXO regulates RNA interference in Drosophila and protects from RNA virus infection. Proc Natl Acad Sci, USA 112, 14587-14592. Stephenson, E.B., Peel, A.J., Reid, S.A., Jansen, C.C., McCallum, H., 2018. The non-human reservoirs of Ross River virus: a systematic review of the evidence. Parasit Vectors 11, 188. Stevens, L.J., Page-McCaw, A., 2012. A secreted MMP is required for reepithelialization during wound healing. Mol Biol Cell 23, 1068-1079. Strauss, J.H., Strauss, E.G., 1994. The alphaviruses: gene expression, replication, and evolution. Microbiol Rev 58, 491-562.

79 Suhrbier, A., La Linn, M., 2004. Clinical and pathologic aspects of arthritis due to Ross River virus and other alphaviruses. Curr Opin Rheumatol 16, 374-379. Sun, P., Nie, K., Zhu, Y., Liu, Y., Wu, P., Liu, Z., Du, S., Fan, H., Chen, C.-H., Zhang, R., 2020. A mosquito salivary protein promotes flavivirus transmission by activation of autophagy. Nat Commun 11, 260. Tabachnick, W.J., 1991. Evolutionary genetics and arthropod-borne disease: the yellow fever mosquito. Am Entomol 37, 14-26. Taft, R.J., Glazov, E.A., Lassmann, T., Hayashizaki, Y., Carninci, P., Mattick, J.S., 2009. Small RNAs derived from snoRNAs. RNA 15, 1233-1240. Taylor, W., Marshall, I., 1975. Adaptation studies with Ross River virus: retention of field level virulence. J Gen Virol 28, 73-83. Tchankouo-Nguetcheu, S., Bourguet, E., Lenormand, P., Rousselle, J.-C., Namane, A., Choumet, V., 2012. Infection by chikungunya virus modulates the expression of several proteins in Aedes aegypti salivary glands. Parasit Vectors 5, 264. Team, R.C., 2013. R: A language and environment for statistical computing. https://www.R- project.org/ (accessed on 18.15.2020). Tenenbaum, D., 2019. KEGGREST: Client-side REST access to KEGG. R package version 1.24. 0. 2019. Timmons, J.A., Szkop, K.J., Gallagher, I.J., 2015. Multiple sources of bias confound functional enrichment analysis of global-omics data. Genome Biol 16, 186. Togami, E., Gyawali, N., Ong, O., Kama, M., Cao-Lormeau, V.-M., Aubry, M., Ko, A.I., Nilles, E.J., Collins-Emerson, J.M., Devine, G.J., 2020. First evidence of concurrent enzootic and endemic transmission of Ross River virus in the absence of marsupial reservoirs in Fiji. Int J Infect Dis 96, 94-96. Tomerini, D.M., 2008. The impact of local government mosquito control programs on Ross River virus disease in Queensland, Australia. Griffith University. Trobaugh, D.W., Klimstra, W.B., 2017. MicroRNA regulation of RNA virus replication and pathogenesis. Trends Mol Med 23, 80-93. Van Rij, R.P., Saleh, M.-C., Berry, B., Foo, C., Houk, A., Antoniewski, C., Andino, R., 2006. The RNA silencing endonuclease Argonaute 2 mediates specific antiviral immunity in Drosophila melanogaster. Genes Dev 20, 2985-2995. Varjak, M., Dietrich, I., Sreenu, V.B., Till, B.E., Merits, A., Kohl, A., Schnettler, E., 2018. Spindle- E acts antivirally against alphaviruses in mosquito cells. Viruses 10, 88. Vazeille, M., Moutailler, S., Coudrier, D., Rousseaux, C., Khun, H., Huerre, M., Thiria, J., Dehecq, J.-S., Fontenille, D., Schuffenecker, I., 2007. Two Chikungunya isolates from the outbreak of La Reunion (Indian Ocean) exhibit different patterns of infection in the mosquito, Aedes albopictus. PLoS One 2, e1168. Vodovar, N., Bronkhorst, A.W., Van Cleef, K.W., Miesen, P., Blanc, H., Van Rij, R.P., Saleh, M.- C., 2012. Arbovirus-derived piRNAs exhibit a ping-pong signature in mosquito cells. PLoS One 7, e30861. Wahid, B., Ali, A., Rafique, S., Idrees, M., 2017. Global expansion of chikungunya virus: mapping the 64-year history. Int J Infect Dis 58, 69-76. Wang, H., Clem, R.J., 2011. The role of IAP antagonist proteins in the core apoptosis pathway of the mosquito disease vector Aedes aegypti. Apoptosis 16, 235-248. Waterhouse, R.M., Kriventseva, E.V., Meister, S., Xi, Z., Alvarez, K.S., Bartholomay, L.C., Barillas- Mury, C., Bian, G., Blandin, S., Christensen, B.M., 2007. Evolutionary dynamics of immune- related genes and pathways in disease-vector mosquitoes. Science 316, 1738-1743. Weaver, S.C., Barrett, A.D., 2004. Transmission cycles, host range, evolution and emergence of arboviral disease. Nat Rev Microbiol 2, 789. Weaver, S.C., Lecuit, M., 2015. Chikungunya virus and the global spread of a mosquito-borne disease. N Engl J Med 372, 1231-1239.

80 Whelan, P.I., Merianos, A., Patel, M., Tai, K., Currie, B., 1993. The epidemiology of arbovirus infection in the Northern Territory 1980-92. https://hdl.handle.net/10137/10281. Wickham, H., 2011. The split-apply-combine strategy for data analysis. J Stat Softw 40, 1-29. Wickham, H., 2016. ggplot2: elegant graphics for data analysis. Springer. Wickham, H., Seidel, D., 2019. scales: Scale Functions for Visualization. https://CRAN.R- project.org/package=scales (accessed on 18.15.2020). Winter, F., Edaye, S., Hüttenhofer, A., Brunel, C., 2007. Anopheles gambiae miRNAs as actors of defence reaction against Plasmodium invasion. Nucleic Acids Res 35, 6953-6962. Xi, Z., Ramirez, J.L., Dimopoulos, G., 2008. The Aedes aegypti toll pathway controls dengue virus infection. PLoS Pathog 4, e1000098. Xia, H., Wang, Y., Atoni, E., Zhang, B., Yuan, Z., 2018. Mosquito-associated viruses in China. Virologl Sinica 33, 5-20. Xu, X., Ji, H., Jin, X., Cheng, Z., Yao, X., Liu, Y., Zhao, Q., Zhang, T., Ruan, J., Bu, W., 2019a. Using pan RNA-Seq analysis to reveal the ubiquitous existence of 5′ and 3′ end small RNAs. Front Genet 10, 105. Xu, X., Ji, H., Jin, X., Cheng, Z., Yao, X., Liu, Y., Zhao, Q., Zhang, T., Ruan, J., Bu, W., 2019b. Using Pan RNA-Seq Analysis to Reveal the Ubiquitous Existence of 5′ and 3′ End Small RNAs. Front Genet 10, 105. Yi, R., Qin, Y., Macara, I.G., Cullen, B.R., 2003. Exportin-5 mediates the nuclear export of pre- microRNAs and short hairpin RNAs. Genes & development 17, 3011-3016. Ylla, G., Fromm, B., Piulachs, M.-D., Belles, X., 2016. The microRNA toolkit of insects. Sci Rep 6, 37736. Zambon, R.A., Vakharia, V.N., Wu, L.P., 2006. RNAi is an antiviral immune response against a dsRNA virus in Drosophila melanogaster. Cell Microbiol 8, 880-889. Zhang, G., Hussain, M., O’Neill, S.L., Asgari, S., 2013. Wolbachia uses a host microRNA to regulate transcripts of a methyltransferase, contributing to dengue virus inhibition in Aedes aegypti. Proc Natl Acad Sci, USA 110, 10276-10281. Zhang, R., Hryc, C.F., Cong, Y., Liu, X., Jakana, J., Gorchakov, R., Baker, M.L., Weaver, S.C., Chiu, W., 2011. 4.4 Å cryo‐EM structure of an enveloped alphavirus Venezuelan equine encephalitis virus. EMBO J 30, 3854-3863. Zhao, C., Li, T., Zhao, Y., Zhang, B., Li, A., Zhao, S., Hou, L., Xia, H., Fan, S., Qiu, J., 2020. Integrated small RNA and mRNA expression profiles reveal miRNAs and their target genes in response to Aspergillus flavus growth in peanut seeds. BMC Plant Biol 20, 1-16. Zhao, L., Alto, B.W., Shin, D., 2019. Transcriptional profile of Aedes aegypti Leucine-Rich Repeat Proteins in response to Zika and Chikungunya viruses. Int J Mol Sci 20, 615. Zhao, L., Alto, B.W., Smartt, C.T., Shin, D., 2018. Transcription profiling for defensins of Aedes aegypti (Diptera: Culicidae) during development and in response to infection with Chikungunya and Zika viruses. J Med Entomol 55, 78-89. Zhao, W., Zhou, G., He, H., 2000. Cloning and primary analysis of 3'end genome of two alphaviruses isolated from Hainan Province of China. Zhonghua Shi Yan He Lin Chuang Bing Du Xue Za Zhi 14, 213-317. Zhou, Z., Li, X., Liu, J., Dong, L., Chen, Q., Liu, J., Kong, H., Zhang, Q., Qi, X., Hou, D., 2015. Honeysuckle-encoded atypical microRNA2911 directly targets influenza A viruses. Cell Res 25, 39-49. Zhu, A., Ibrahim, J.G., Love, M.I., 2019. Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences. Bioinformatics 35, 2084-2092.