Characterization of Autographa californica nucleopolyhedrovirus immediate early protein ME53: The role of conserved domains in BV production, viral gene transcription, and evidence for ME53 presence at the ribosome

by

Robyn Ralph

A Thesis presented to The University of Guelph

In partial fulfilment of requirements for the degree of Master of Science in Molecular and Cellular Biology

Guelph, Ontario, Canada

© Robyn Ralph, December, 2018

ABSTRACT

CHARACTERIZATION OF AUTOGRAPHA CALIFORNICA NUCLEOPOLYHEDROVIRUS IMMEDIATE EARLY PROTEIN ME53: THE ROLE OF CONSERVED DOMAINS IN BV PRODUCTION, VIRAL GENE TRANSCRIPTION, AND EVIDENCE FOR ME53 PRESENCE AT THE RIBOSOME

Robyn Ralph Advisors: University of Guelph, 2018 Dr. Peter Krell Dr. Sarah Wooton

The baculovirus AcMNPV early/late gene me53 is required for efficient BV production and is conserved in all alpha and betabaculoviruses. The 449-amino acid protein contains several highly conserved functionally important domains including two putative C4 zinc finger domains (ZnF-N and ZnF-C) whose cysteine residues are 100% conserved. One purpose of this study is to confirm the presence of two zinc binding domains in ME53, as well as determine their role in virus infection and viral gene transcription. Interestingly, deletion of ZnF-C results in an early delay of BV production from 12 to 18 hours post transfection correlating to ME53's cytoplasmic localization.

Cytoplasmic functions at early times post-transfection may include translational regulation, which is supported by yeast-2-hybrid data that ME53 interacts with the host

40S ribosomal subunit protein RACK1. In this study the association of ME53 with the ribosomes of virus infected cells was also investigated. iii

DEDICATION

I dedicate this thesis to my father, Ronald James Ralph. Through him I learned that in life there are always choices. Choices to learn, choices to succeed, and choices to love. The tenacity he showed throughout his life to ensure that his time was well spent in his favourite places with the people he loved, is a quality I hope to emulate. Time is the most valuable possession and I am blessed to have learned that lesson early in life. To my dad, I love you.

In the Big Rock Candy Mountains, All the cops have wooden legs And the bulldogs all have rubber teeth And the hens lay soft-boiled eggs. The farmers' trees are full of fruit And the barns are full of hay, Oh I'm bound to go Where there ain't no snow Where the rain don't fall The wind don't blow In the Big Rock Candy Mountains. - “Big Rock Candy Mountain”, Harry McClintock

iv

ACKNOWLEDGEMENTS The research I completed during this thesis would not have been possible without the guidance and support of Dr. Peter J. Krell. Thank you for introducing me to virology, and providing your expertise throughout my undergraduate and Master’s degrees, it has been an invaluable experience.

I would also like to acknowledge Dr. Sarah Wootton, for the advice and valuable discussions during my research and thesis writing. In addition, I would like to thank Dr.

Baozhong Meng and Dr. Steffen Graether for the support and suggestions that kept me focused and on track.

Past and present members of the Krell and Meng labs provided endless help throughout my research. To Dr. Emine Ozsahin, thank you for always having time for my questions, and for the friendship that made long days and nights in the lab all that more enjoyable. Thank you as well to Sunny for the friendship and support, especially towards the end of my research where you were vital to the completion of several experiments.

To Dr. Huogen Xiao, thank you for the guidance and help, and Clayton Moore for the friendship and laughs through two and a half years.

Finally, I need to thank my family and friends who were my cheerleaders through thick and thin. Specifically, my mother Anita for your love and patience while I wrote in your living room, and my husband Josh for taking everything in stride and never failing to support me. As this chapter closes, I look forward to where we will go.

v

TABLE OF CONTENTS

ABSTRACT ...... II

DEDICATION ...... III

ACKNOWLEDGEMENTS ...... IV

TABLE OF CONTENTS ...... V

LIST OF TABLES ...... VIII

LIST OF FIGURES ...... IX

ABBREVIATIONS ...... XI

CHAPTER 1: LITERATURE REVIEW ...... 1

Baculoviruses ...... 1

Baculovirus virion phenotypes ...... 3

Viral life cycle ...... 3

Immediate early gene expression regulation and function ...... 6

Viral DNA replication and late gene expression ...... 10

Baculovirus strategies to evade the host anti-viral response ...... 11

AcMNPV ME53 ...... 15

ME53 intracellular localization...... 19

The association of ME53 with the capsid and its role in DNA replication...... 21

ME53 and the transcription of viral genes ...... 22

C4 zinc finger proteins ...... 23

Thesis Objectives ...... 26

vi

CHAPTER 2: BIOINFORMATIC ANALYSIS OF ME53 SECONDARY STRUCTURE AND CONFIRMATION OF ZINC FINGER DOMAINS, AND ME53 OVEREXPRESSION 29

Introduction ...... 29

Methods ...... 31 2.2.1 Bioinformatic analysis of ME53 disorder, binding prediction, and redox potential ...... 31 2.2.2 Cell lines ...... 32 2.2.3 Peptide resuspension ...... 32 2.2.4 Circular dichroism ...... 34 2.2.5 Analysis of CD data ...... 34 2.2.6 Construction of ME53 overexpression bacmid ...... 35 2.2.7 Bacmid DNA isolation ...... 36 2.2.8 Virus amplification ...... 36 2.2.9 Estimation of viral titre using the end-point dilution assay ...... 37 2.2.10 Multiplicity of infection calculation...... 38 2.2.11 His-ME53 purification ...... 38 2.2.12 SDS-PAGE and Western blot analysis...... 39

Results ...... 40 2.3.1 Structural Analysis of ME53 ...... 40 2.3.2 Circular dichroism analysis of ME53 peptides ...... 46 2.3.3 Construction of His-ME53 overexpression bacmid ...... 48 2.3.4 Purification of His-ME53 ...... 50

Discussion ...... 54

CHAPTER 3: THE ROLE OF ME53 DOMAINS IN BV PRODUCTION AND VIRAL GENE TRANSCRIPTION ...... 60

Introduction ...... 60

Methods ...... 62 3.2.1 Bioinformatic analysis of ME53 ...... 62 3.2.2 ME53 knockout bacmid construction ...... 63 3.2.3 Transfections for growth curve analysis ...... 67 3.2.4 Determination of viral titre...... 68 3.2.5 Viral gene selection for qRT-PCR analysis ...... 71 3.2.6 Transfection for transcriptional analysis ...... 72 3.2.7 RNA isolation from transfected cells ...... 72 3.2.8 Generation of cDNA using reverse transcription ...... 73 3.2.9 Primer amplification efficiency assay ...... 73 3.2.10 Viral transcript analysis using qRT-PCR ...... 75

Results ...... 75

vii

3.3.1 Identification of conserved regions targeted for deletion/mutagenesis ...... 75 3.3.2 Generation of bacmids with ME53 deletion/mutation variants ...... 80 3.3.3 Virus replication of AcMNPV with conserved domain deletions in me53 ...... 80 3.3.4 The role of zinc coordination by ME53 in BV production ...... 85 3.3.5 Selection of viral genes for gene expression analysis ...... 88 3.3.6 Primer amplification efficiency verification ...... 91 3.3.7 Transcription analysis ...... 93

Discussion ...... 103

CHAPTER 4: ME53 ASSOCIATES WITH THE RIBOSOME IN VIRUS INFECTED CELLS ALTHOUGH RACK1 PRESENCE AT THE RIBOSOME IS DEPLETED ...... 113

Introduction ...... 113

Methods ...... 115 4.2.1 Ac+GFP:ME53 recombinant bacmid construction ...... 115 4.2.2 Bacmid DNA isolation ...... 117 4.2.3 Virus amplification and titration ...... 117 4.2.4 Polysome Profiling ...... 118 4.2.5 Trichloracetic acid (TCA) Precipitation ...... 119 4.2.6 SDS-PAGE and Western blot analysis ...... 119

Results ...... 120 4.3.1 GFP tagged recombinant bacmid construction and virus amplification ...... 120 4.3.2 AcMNPV protein ME53 associates with ribosomes in baculovirus infected cells 123 4.3.3 Bacluovirus infected cells show decreased concentration of RACK1 at the ribosome than uninfected cells...... 126

Discussion ...... 129

CHAPTER 5: GENERAL CONCLUSIONS AND FUTURE DIRECTIONS ...... 133

REFERENCES ...... 139

viii

LIST OF TABLES

Table 2.1 ME53 Zinc Finger Peptides ...... 33

Table 3.1: Baculoviruses and their abbreviations and accession numbers used in the ME53 phylogenetic analysis ...... 64

Table 3.2: Primers for internal deletions and mutations of me53 in pFACTproME53:GFP. All primers are 5’-phosphorylated...... 69

Table 3.3: Primer pair sequences used in the qRT-PCR assay for viral gene expression analysis...... 74

Table 3.4: Tetranucleotide promoter consensus motifs identified upstream of the ORF with the “A” of the ATG start codon representing +1...... 90

Table 3.5 Sequence identity matrix of the promoter sequences 300 bp upstream of the translation start site of viral genes selected for qRT-PCR analysis...... 92

ix

LIST OF FIGURES

Figure 1.1: The four clades of Baculoviridae...... 2 Figure 1.2: Diagram of baculovirus virion structures...... 5 Figure 1.3: Interconnectivity networks of immediate early genes...... 9 Figure 1.4: The eukaryotic 40S ribosomal subunit (PDB ID 2XZM)...... 16 Figure 1.5: Alignment of ME53 sequences ...... 18 Figure 1.6: Transcription factor IIIA from Xenopus laevis ...... 25 Figure 2.1: IUPred2A analysis of ME53...... 41 Figure 2.2 JNet Pred Output for ME53...... 44 Figure 2.3 Solution CD spectra analysis by CAPITO of the ZnF-N peptide in the presence and absence of zinc...... 48 Figure 2.4 CD spectra analysis by CAPITO of the ZnF-C peptide in 20 mM Tris, pH 7.2, in the presence and absence of zinc...... 47 Figure 2.5: Comparison of the secondary structure prediction similarities of AcMNPV ME53 ZnF-C and Ubiquitin ...... 49 Figure 2.6: Map of his-tagged ME53 insertion into the commercial bacmid bMON14272 genome...... 51

Figure 2.7: TCID50 titre of GFP-only and His-tagged ME53 viruses...... 52 Figure 2.8: Validation of HisME53 purification...... 53 Figure 3.1: Maximum likelihood phylogenetic tree of ME53 sequences...... 78 Figure 3.3.2: Alignment of ME53 sequences...... 79 Figure 3.3: Schematics of bacmids generated for growth curve and RT-qPCR analysis...... 81 Figure 3.4 The effect of ME53 domain deletions on viral growth kinetics...... 84 Figure 3.5: The effect of ME53 zinc coordination on BV production...... 87 Figure 3.6: Analysis of primer amplification efficiencies for RT-qPCR...... 94 Figure 3.7: Relative transcript levels of ie1...... 96 Figure 3.8: Relative transcript levels of lef-8...... 97 Figure 3.9: Relative transcript levels of lef-9...... 98 Figure 3.10: Relative transcript levels of vp80...... 100 Figure 3.11: Relative transcript levels of gp64...... 101 Figure 3.12: Relative transcript levels of odv-e25...... 102 Figure 4.1 Construction of GFP:ME53 bacmid ...... 121

x

Figure 4.2 Titre of GFP-only and GFP-tagged ME53 viruses using the end point dilution assay method...... 122 Figure 4.3 Polysome profilies of uninfected and GFP:ME53 virus infected cells at 12 and 20 hpi...... 125 Figure 4.4: Total concentration of RACK1 in uninfected and infected cellsm and the localization of RACK1 at the ribosome...... 127

xi

ABBREVIATIONS aa Amino acid AcMNPV Autographa californica nucelopolyhedrovirus BEVS Baculovirus expression vector system BIR Baculovirus IAP repeat BmNPV Bombyx mori nucleopolyhedrovirus bps Base pairs BV Budded virus CAPITO CD Analysis and Plotting Tool cat Chloramphenicol acetyl transferase CD Circular dichroism CPE Cytopathic effects eIF2a Eukaryotic initiation factor 2a FBS Fetal bovine serum GCN2s General control nonrepressible-2-kinases GFP Green fluorescent protein GV Granulovirus HCV Hepatitis C Virus HearNPV Helicoverpa armigera nucleopolyhedrovirus His Histidine tag HP Heavy polysome hpi Hours post infection hpt Hours post transfection HRI-like Heme-regulated inhibitor kinases HRP Horseradish peroxidase hrs Homologous regions IAPs Inhibitors of apoptosis IRES Internal ribosomal entry sites L Linker region LEF Late expression factor LP Light polysome MAPK Mitogen activated protein kinase mdeg Millidegrees mqH2O Milli Q H2O MOI Multiplicity of Infection NPV Nucleopolyhedrovirus NTS Nuclear translocation sequence OBs Occlusion bodies ODV Occlusion-derived virus

xii

PBS Phosphate buffered saline PDB Protein Data Bank PERK PKR-like endoplasmic reticulum kinases PIFs Per os infectivity factors PKCβII Protein kinase C βII PKR Protein kinase R PMSF Phenylmethanesulfonyl PTMs Post-translational modifications PVDF Polyvinylidene difluoride RACK1 Receptor for activated protein kinase C βII RING finger Really interesting new gene finger RNP Ribonucleoprotein SFM Serum Free Media TBS Tris buffered saline TBS-T Tris buffered saline with Tween TCA Trichloracetic acid TFIIIA Transcription factor IIIA WT Wild-type ZnF-C C-terminal zinc finger ZnF-N N-terminal zinc finger

1

Chapter 1: Literature review

Baculoviruses

Baculoviruses are entomopathogenic viruses of the Baculoviridae family infecting insect hosts in the orders Diptera, Hymenoptera, and Lepidoptera of the Insecta class

(Jehle et al., 2006). They are characterized by enveloped, rod-shaped virions with double- stranded, circular, supercoiled genomes, which vary in size from 80-180 kb (Rohrmann

2013). The Lepidopteran order contains the majority of known permissive species (90%), and lepidopteran-specific baculoviruses separate into two genera, Alphabaculovirus, and

Betabaculovirus that consist of the lepidopteran nucleopolyhedroviruses (NPVs), and granuloviruses (GVs) respectively (Jehle et al., 2006). The Baculoviridae contains two other genera, Gammabaculovirus consisting of hymenopteran NPVs and

Deltabaculovirus for the dipteran NPVs (Figure 1.1) (Rohrmann, 2013).

Alphabaculoviruses are further classified into Group I and Group II NPVs and are distinguished by their different fusogenic proteins (Miele et al., 2011). Group I NPVs use

GP64 and Group II GVs, lacking GP64, use F proteins instead for cell entry (Rohrmann,

2013). The different cell entry proteins are the result of an evolutionary divergence in alphabaculoviruses that also generated 11 unique open reading frames exclusive to

Group I alphabaculoviruses (Miele et al., 2011). Autographa californica multiple nucleopolyhedrovirus (AcMNPV), the virus used in this study, is a Group I alphabaculovirus that infects larvae of several Lepidopteran species; although, the virus is named for the Alfalfa looper moth (Rohrmann, 2013). AcMNPV has a broader host range relative to other baculovirus species, which are typically specific to one, or only a few, host species, and is the predominant focus of this literature review.

2

Figure 1.1: The four clades of Baculoviridae. Phylogenetic tree of 29 core proteins found in the 4 genera of the Baculoviridae family represented by the four clades. The alphabaculoviruses are shown separated into Group I and Group II with alphabaculovirus AcMNPV (circled) part of Group I. This image is adapted from Jehle et al., 2006.

3

Baculovirus virion phenotypes

One trademark of baculoviruses is their biphasic replication cycle that produces two structurally distinct virion phenotypes, termed budded virus (BV) and occlusion- derived virus (ODV) (Blissard, 1996). BVs are produced during the late phase of infection via budding of a single nucleocapsid through the infected cell’s plasma membrane, and they facilitate cell to cell transmission of the virus within the host (or cell culture) (Blissard and Rohrmann, 1990). Occlusion bodies (OBs) contain membrane bound occlusion- derived virions (ODVs) that are occluded by a crystalline viral polyhedrin protein matrix as individual or multiple nucleocapsids in the nucleus during late infection forming polyhedra (Blissard, 1996). Occlusion bodies are capable of remaining viable in harsh environments and enable horizontal transmission of the virus between hosts via oral ingestion (Hou et al., 2012). BVs and ODVs are genomically identical and structurally similar; however, their envelopes and associated structures differ to serve separate functional purposes (Figure 1.2) (Jehle et al., 2006).

Viral life cycle

Baculovirus infection, including that of AcMNPV, begins with ingestion of OBs from the environment by the host and their subsequent dissolution in the alkaline environment of the lepidopteran midgut (Lauzon et al., 2004). Dissolution of the crystalline OB releases ODVs into the lumen and ODVs must subsequently pass through a protective peritrophic membrane to fuse with the midgut microvillar membranes of columnar epithelial cells (Lauzon et al., 2004). Successful infection of the midgut epithelial cells relies on ODV membrane components termed per os infectivity factors (PIFs) (Peng et al., 2010). Deletion of any of the five PIF genes, pif-1, pif-2, pif-3,

4 pif-4, and p74 render ODVs non-infectious, and PIF-1, PIF-2, and PIF-3 form a complex involved in receptor-mediated endocytosis (Peng et al., 2010). P74 also interacts with this complex but is not required for nucleocapsid entry and therefore may be involved in nucleocapsid transport or mediation of cell-signaling pathways (Peng et al., 2010). PIF-

4 is associated with the envelopes of both virion phenotypes and is essential for BV infectivity and ODV morphology (Tao et al., 2013).

PIF-mediated fusion of ODVs with the mature columnar midgut epithelial cells results in nucleocapsid entry into the cell (Peng et al., 2010). Nucleocapsids containing viral genetic information are transported along actin filaments and through nuclear pores into the nucleus (Blissard and Rohrmann, 1990). Viral DNA is subsequently uncoated and immediate early genes including ie1, ie0, ie2, pe38, and me53 are expressed (Ono et al.,

2015). Immediate early genes are required for the initiation of viral infection as they transactivate delayed immediate-early genes such as DNA pol, lef-3, and lef-4. The cognate gene products stimulate viral DNA replication as early as 6 hours post infection

(hpi) (Lu et al., 1997). Viral DNA replication occurs in the nucleus from 6 – 18 hpi and is followed by viral RNA polymerase-dependent late gene expression of structural proteins involved in virus production (Rohrmann, 2013). Progeny nucleocapsid assembly begins around 12 hpi, and the nucleocapsids are trafficked to the plasma membrane for budding during late infection (Blissard, 1996). BV nucleocapsids and ODVs are both assembled in the nucleus, but the timing of their production is separated by the late and very late phases of infection. The AcMNPV life cycle is dependent on strict temporal regulation of viral gene expression and the orderly pattern of viral DNA replication, assembly, and virion transport. Temporal regulation is important for the evasion of host defense systems,

5

A B C

Figure 1.2: Diagram of baculovirus virion structures. (A) OBs contain one or several nucleocapsids enveloped and embedded in a crystalline polyhedrin matrix. (B) The ODV contains several nucleocapsids wrapped in a nuclear derived envelope. (C) Nucleocapsids found in budded virions are the same as those in ODVs but their membranes are derived from the plasma membrane and contain the viral glycoprotein GP64 or F protein depending on their taxonomic classification. Figure from Au et al., 2013.

6 to reduce redundancy in the genome, and to maximize viral production (Theilmann and

Blissard, 2008).

Immediate early gene expression regulation and function

Transcription of baculovirus genes occurs over four temporal phases that rely on gene products from the previous phase to activate a regulatory cascade of transcription

(Rohrmann, 2013). The four phases are: immediate early, delayed early, late, and very late. The immediate early and early phases utilize host RNA polymerase II through TATA elements that resemble host transcription initiation sites, and other cis-acting regulatory elements such as the CAGT initiator motif (Jiang et al., 2006; Bleckmann et al., 2015).

Several highly conserved immediate early gene products are transactivating proteins, such as IE-1, that stimulate the transcription of early genes while total viral DNA concentration is low (Jiang et al., 2006, Rohrmann, 2013). Other immediate early genes such as me53, the focus of this thesis, are highly conserved but their immediate early functions during infection remain to be determined (Knebel-Mörsdorf et al., 1993, Knebel-

Mörsdorf et al., 1996).

Products of five immediate-early genes, ie-0, ie-1, ie-2, pe38, and me53 have confirmed or suggested transcriptional regulatory functions. IE-1 and IE-0 are splice variants that universally activate the transcription of viral genes and enhance the expression of late expression factors, respectively (Kovacs et al., 1993; Theilmann et al.,

2001). IE-2 contains a predicted RING (Really Interesting New Gene) finger domain and is involved in the indirect transactivation of viral genes, the regulation of cell cycle arrest, and E3 ubiquitin ligase activity while enhancing the expression of heat shock proteins for increased stability of IE-2 (Prikhod’ko and Miller, 1998; Imai et al., 2005; Tung et al.,

7

2016). PE38 also contains a predicted RING finger motif that is involved in E3 ubiquitin ligase activity and enhances apoptosis in insect cells when transiently expressed (Imai et al., 2003; Prikhod’ko and Miller, 1998), it is also essential for viral DNA replication as its deletion results in an attenuation of viral DNA production in vivo (Milks et al., 2003). The fifth immediate early gene product, me53, is expressed as early as 30 min post infection and is essential for optimal BV production (Knebel-Morsdorf et al., 1993; de Jong et al.,

2009). ME53 contains two putative C4 zinc finger domains and an identified nuclear translocation sequence (NTS) that facilitates its nuclear localization at late times post infection (Knebel-Morsdorf et al., 1993; Liu et al., 2016). ME53 is implicated in the upregulation of late gene expression through an as of yet unknown mechanism (Liu,

2015).

The transactivating effects of the aforementioned immediate-early genes have been investigated in predominately transient or pairwise expression assays or through reverse genetic analyses. Transient expression analyses examining the effect of IE proteins on the IE gene promoters using luciferase expression assays found that PE38 activates solely the ie-1 promoter. The IE-1 splice variant IE-0 is self-repressive, but its promoter is activated by IE-1, IE-2, ME53, and PE38 (Ono et al., 2015). IE2 is cis-acting, and its promoter is activated by IE1 but repressed by ME53 (Ono et al., 2015). PE38 was similar to IE-0 in that it is negatively auto-regulated, but the four other immediate early proteins activate the pe38 promoter. ME53 is the only immediate early protein to have a negative outgoing link in the transient expression interaction network through its repression of the ie-2 promoter (Figure 1.3). The reduced activity of the ie2 promoter in the presence of

ME53 suggests that ME53 functions as a negative temporal regulator (Ono et al., 2015).

8

Reverse genetic analyses using qRT-PCR to monitor IE gene expression at 6 hpt supported fewer than half of the interactions suggested in the transient expression analysis. Specifically, deletion of ie-1 from the bacmid genome results in the downregulation of the four other immediate early genes consistent with its role as a general transactivator. The deletion of ie-0 and ie-2 from the genome results in an increase in the expression of other immediate early genes suggesting that ie-0 and ie-2 act as temporal repressors. ME53 and PE38 have no outgoing regulatory links indicating that deletion of their genes did not affect the transcription of other immediate early genes when deleted.

The discrepancies between the two sets of results, i.e. the upregulation of the ie-1 and ie-0 promoters by the transient expression of PE38 but the absence of PE38- dependent upregulation of ie-1 and ie-0 transcripts in the reverse genetic analyses, show that immediate early gene functions depend on the presence of other viral proteins. The difference in analyitical methods used, protein expression vs. transcript levels also provided different interpretations. For luciferase expression in the transient expression assays, both transcription and translation must occur while transcription only is needed for the transcript level analysis in the reverse genetic approach. In addition, some immediate early genes may have a role in viral transcript processing, such as mRNA processing or transport into the cytoplasm, or may interact with translation initiation factors and ribosomes for preferential translation of viral transcripts. Also, the reverse genetics analysis was performed at 6 hpt and some immediate early proteins, like ME53, do not localize to the nucleus until later times post-transfection, minimizing their potential to affect the immediate early gene transcript levels (Liu et al., 2016). The apparently

9

A B

Figure 1.3: Interconnectivity networks of immediate early genes. Interconnectivity network of several immediate early genes using a transient expression assay (A) and reverse genetics (B) with blue arrows representing an activating function and red lines indicating a repressive function. A) Interaction network of immediate early genes measured by transiently expressing each immediate early gene and quantifying their effect on the expression levels of luciferase under the respective immediate early promoters. B) Network of interactions between immediate early genes by deleting each immediate early gene individually and quantifying the mRNA of the remaining immediate early genes. Data interpreted from Ono et al., 2015.

10 contradictory results suggest that immediate early proteins have complex interactions with each other, other viral proteins, or host proteins to directly, or indirectly regulate the expression of other immediate early genes during infection.

Viral DNA replication and late gene expression

The AcMNPV genome contains homologous regions (hrs) found at seven locations around the circular genome that each contain different numbers of copies of a 30-bp palindrome near their center (Vanarsdall et al., 2007). The homologous regions are thought to be both transcriptional enhancers and possible origins of DNA replication with

IE1 binding to them in AcMNPV to enhance transcription. The timing of viral DNA replication initiation temporally separates the early and late phases of transcription and is initiated by delayed early gene products such as LEF-1 and LEF-2 (Mikhailov and

Rohrmann, 2002). Baculoviruses use promoters to control the temporal expression of genes and will halt transcription from early promoters, through a process not completely understood, once DNA replication begins (Rohrmann, 2013). Following viral DNA replication, viral RNA polymerase produces a high number of late and very-late transcripts for genes whose proteins are associated with virion structure, DNA packaging into nucleocapsids, and nucleocapsid transport to the plasma membrane (Rohrmann, 2013).

Baculoviruses hijack the host transcriptional machinery for expression of early viral genes but rely on their own RNA polymerase for late gene transcription after viral DNA replication initiation (Acharya and Gopinathan, 2002). Late transcription uses a virally encoded RNA polymerase made up of several core late expression factor (LEF) gene products conserved across baculoviruses (LEF-8, LEF-9, LEF-4, and P47) (Guarino et al., 1998). The late phase of transcription involves roughly half of the baculovirus genes,

11 encoding mostly viral structural proteins or proteins involved in nucleocapsid assembly

(Beniya et al., 1996). The majority of the late genes expressed are under the influence of an RTAAG sequence motif (Jiang et al., 2006). There is a higher amount of late viral transcripts relative to early transcripts due to viral genomic amplification and the high activity of the viral RNA polymerase (Jiang et al., 2006). The immediate early protein, IE0, also influences late gene transcription suggesting that other early gene products may regulate transcription after DNA replication (Passarelli and Guarino, 2007).

Baculovirus strategies to evade the host anti-viral response

Viruses rely on their hosts’ systems for progeny virus production and proliferation. For example, baculoviruses can initiate replication by only transfection of the viral genome without any viral proteins. Infection is accomplished by cis-acting viral gene promoter consensus sequences that exploit the host RNA polymerase II for immediate early gene transcription. The immediate early proteins then initiate a cascade of viral protein expression by acting as transactivators for genes involved in viral DNA replication and virion production. An essential component of viral infection is the evasion or repression of host immune responses to optimize virus production. Insects lack the adaptive immune response present in other eukaryotes and rely on multiple physical barriers and an innate immune response for defence against pathogens (Ikeda et al., 2013).

Two dominant anti-viral pathways in the innate immune response are apoptosis and global protein synthesis shutdown (Du and Thiem, 1997b). Baculoviruses circumvent these defenses using several mechanisms, including the adaptation of genes acquired from the host throughout evolution to inhibit the activation of these pathways (Ikeda et al.,

2013). The baculovirus P35 protein was the first protein ever shown to have antiapoptotic

12 activity during baculovirus infection. P35 is presumed to be derived from a host gene and prevents apoptosis in baculovirus-infected cells by inhibiting group I, II, and III caspases

(Lin et al., 2001). P35 contains a caspase cleavage motif, DQMD, that is recognized by the host effector caspases (Clarke and Clem, 2003). Upon cleavage, the two P35 fragments become permanently associated with the effector caspases and form a stable complex that has no caspase activity (Clarke and Clem, 2003). In addition, cell lines transiently expressing P35 are resistant to nutrient depletion and other canonical stresses that would induce apoptosis (Lin et al., 2001).

Baculoviruses also encode inhibitors of apoptosis (IAPs) that are characterized by baculovirus IAP repeat (BIR) domains located at the N-terminal region, and a RING finger domain found at the C-terminus (Mace et al., 2010). Baculovirus IAPs have cellular homologues in yeast, mammals, and other organisms whose functions range from suppression of apoptosis to differentiation (Ikeda et al., 2013). The baculovirus IAPs are classified into five groups, iap1 to iap5, with iap1-4 found in alpha- and betabaculoviruses

(Clem, 2007). Several IAPs from each group are encoded by individual baculoviruses, but only some IAPs have documented anti-apoptotic activity. Their function appears to be dependent on the combination of cell-type and virus, but many questions remain regarding their functions (Ikeda et al., 2013).

Baculovirus infection-induced apoptosis correlates with the onset of DNA replication when cellular checkpoints identify viral DNA as aberrant or damaged DNA and prompt apoptotic signaling (Wu et al., 2013). Interestingly, global translation arrest is also correlated with the onset of viral DNA replication and is enhanced by apoptotic suppression. Ld652Y cells, derived from the gypsy moth, are non-permissive to AcMNPV

13 because apoptotic suppression by AcMNPV p35 and other IAP proteins causes global translation arrest in these cells (Thiem and Chejanovsky, 2004). Deletion of p35 from

AcMNPV improves the translational activity of Ld652Y cells but does not render these cells permissive for AcMNPV. The host range gene, hrf-1, encoded by Lymantria dispar multiple nucleopolyhedrovirus (LdMNPV), precludes global translation shut off when expressed by AcMNPV and renders the Ld652Y cell line permissive to AcMNPV (Thiem et al., 1996; Du and Thiem, 1997a). The function of Hrf-1 in the regulation of protein synthesis indicates that AcMNPV likely encodes an inhibitor of translation arrest specific to its permissive host species.

The mechanisms to initiate apoptosis and global translation arrest during baculovirus infection are not well characterized, but the correlation between apoptotic suppression and translation arrest suggests that the initiation of these anti-viral defence pathways have similar triggers. The predominant instigator of apoptosis and translation arrest is viral DNA replication; however, the early gene activity of IE1, IE2, and PE38 can result in apoptosis before viral DNA replication and translational inhibition, which occur as early as 8 hpi (Thiem and Chejanovsky, 2004; Schultz et al., 2009). Transient expression of

IE1 in Sf21 cells results in apoptosis, and when expressed in conjunction with PE38, the rate of apoptosis increases by two-fold (Prikhod’ko and Miller, 1999). Transient expression of PE38 does not result in apoptosis and therefore may enhance the apoptotic activity of IE1 in Sf21 cells (Prikhod’ko and Miller, 1999). Separate from its transactivating activities, IE2 results in cell cycle arrest in the S phase via its RING finger domain resulting in abnormally high DNA concentrations, a known apoptotic trigger (Prikhod’ko and Miller,

1998). The ability of immediate early proteins to induce apoptosis in the absence of viral

14 infection implies that other immediate early proteins must act as apoptotic suppressors, or perhaps translational shutdown inhibitors to offset the apoptotic inducing activities of

IE-1, IE2, and PE38. One such possible protein is ME53, an immediate early protein has two putative C4 zinc finger domains and is highly conserved in all alpha and betabaculoviruses and is the focus of this thesis.

The use of baculoviruses in recombinant protein production systems which exploits the mass production of very late proteins p10 and polyhedrin, suggests that the regulation of pathways involved in proteotoxic stress and translational regulation are crucial for baculovirus infection. Yeast-2-hybrid studies completed by Dr. Emine Ozsahin, a post- doctoral fellow, suggest an interaction between the immediate early protein ME53 and the ribosomal protein RACK1 (personal communication). This interaction implicates

ME53 in a translational role during baculovirus infection, specifically during early times when ME53 is predominantly localized in the cytoplasm. The host scaffolding protein

RACK1 is a WD-repeat protein that was first identified as the receptor for activated protein kinase C βII (PKCβII), but is now known to interact with a variety of other proteins (Adams et al., 2011). RACK1 binds at the head of the 40S ribosomal subunit at the mRNA exit channel and associates with other ribosomal proteins (Figure 1.4). RACK1 aids in translational initiation by facilitating phosphorylation of translation initiation factors and acts as a signaling hub at the ribosome (Rabl et al., 2011). Several viruses require RACK1 for virus infection, including Hepatitis C Virus (HCV), which requires RACK1 for the translation of transcripts with internal ribosomal entry sites (IRES), and poxviruses, which customize RACK1 through phosphorylation to preferentially translate viral transcripts with extended polyA leaders (Majzoub et al., 2014; Jha et al., 2017).

15

During periods of proteotoxic stress, RACK1 is sequestered into stress granules and prevents the assembly of the translation initiation complex thereby suppressing translation (Arimoto et al., 2008). The mitogen activated protein kinase (MAPK) p38b is required for RACK1 dissociation from the 40S subunit of the ribosome during proteotoxic stress (Belozerov et al., 2014). RACK1 is phosphorylated by p38b resulting in its sequestration into a ribosome-unbound pool that prevents the coordination of PKC by RACK1 at the ribosome for translation initiation. Some viruses, such as the Group I alphabaculovirus Bombyx mori nucleopolyhedrovirus (BmNPV), exploit the MAPK pathways for virus infection (Katsuma et al., 2007). Interestingly, inhibition of the p38

MAPK pathway does not affect baculovirus production suggesting that it is unnecessary or is already inhibited during infection. The involvement of p38 MAPKs in other viral infections indicates that the latter is more likely, but no baculovirus proteins have been identified to bind P38 directly. The interaction of ME53 with RACK1 may implicate ME53 in a translational regulatory role and is supported by the role of P38 in RACK1 dissociation from the ribosome and BmNPV infection.

AcMNPV ME53

The baculovirus AcMNPV genome (NC_001623.1) consists of 133,894 base pairs

(bps) with an estimated 156 open reading frames with protein-coding potential (Ayers et al., 2004; Jehle et al., 2006). One of the AcMNPV genes, me53, is highly conserved and found in all lepidopteran baculoviruses sequenced to date (Liu, 2015). The AcMNPV

ME53 polypeptide is 53 kDa in size and 449 amino acids (aa) in length. It contains an identified nuclear translocation sequence (NTS) in a highly conserved region of the N- terminus from residues 109-137, two putative C4 zinc finger domains, the N-terminal zinc

16

A

150°

B

Figure 1.4: The eukaryotic 40S ribosomal subunit (PDB ID 2XZM). The eukaryotic 40S ribosomal subunit (PDB ID 2XZM) with the ribosomal proteins shown as surface representation in light grey, and the rRNA shown in dark grey. (A) RACK1 is shown in ribbon form and highlighted by the red arrow at the head of the ribosome near the mRNA exit channel shown within the red circle. The ribosomal proteins that it interacts with are shown as surface representations in yellow, green, and red. (B) RACK1 shown in ribbon representation with its ribosomal protein binding partners, rps3e is shown in green, rps16e in yellow, and rps17e in red respectively.

17 finger (ZnF-N, residues 170-209) and the C-terminal zinc finger (ZnF-C, resiudes 379-

399), and a highly conserved region from residues 278-302, here designated the linker or L region, which may act as a flexible linker between the zinc fingers that stiffens upon ligand binding (Ralph, unpublished results; Knebel-Mörsdorf et al., 1993; Liu et al., 2016).

The cysteine residues of the putative C4 zinc fingers are 100% conserved in all alpha and betabaculoviruses (Figure 1.5).

The me53 promoter has a dual early/late function and forms a divergent transcriptional unit with the IE0 promoter that is transcriptionally active from 30 minutes pi up to 72 hpi (Knebel-Morsdorf et al., 1996; de Jong et al., 2009). Deletion of the early promoter motif CACAGT results in a 360-fold decrease in BV production while deletion of the late promoter motif ATAAG, and both promoters, results in 1,000- and 5,000-fold reductions compared to wild-type virus levels, respectively (de Jong et al., 2009). The late promoter has a more significant effect on virus production than the early promoter, suggesting that the function of ME53 in late times is more critical for efficient baculovirus infection and BV production in cell culture (de Jong et al., 2009). However, the early promoter is active from 30 min to 24 hpi and the late promoter from 12-72 hpi (Knebel-

Mörsdorf et al., 1996). In this regard, the extended activity of the late promoter could rescue the effect of early promoter deletion on virus production and may explain the disparity in BV production between deletions of the early and late promoters respectively.

ME53 is essential for optimal BV production as me53-null viruses demonstrate significantly smaller plaque sizes than wild-type (WT) virus indicating decreased viral spread (de Jong et al., 2009). When titred at 24h intervals, AcMNPV lacking me53 does not produce a detectable virus titre until 48 hours post transfection and is 3 to 4 logs lower

18

NTS

ZnF-N

L

ZnF-C

Figure 1.5: Alignment of ME53 sequences Alignment of ME53 sequences from a group I alphabaculovirus (AcMNPV), two group II alphabaculoviruses (TnSNPV and LdMNPV), and two betabaculoviruses (PxGV and CpGV). The boxed areas correspond to the NTS (blue), the putative ZnF-N (red), the L region (purple), and the putative ZnF-C (green). Clustal Omega was used for the sequence alignment with default parameters and the graphic was generated by ESPript.

19 when compared to WT virus (de Jong et al., 2009). Infections of cell culture with baculoviruses containing varying truncations of ME53 determined that residues 101-398 are required for optimal BV production (de Jong, 2011).

Further work by Liu et al., 2016 using internal deletions of ME53 identified residues

109-137 as essential for virus production and is the region required for nuclear translocation of ME53 during infection. Interestingly, an internal deletion including a segment (aa 160-191) of the recently discovered putative ZnF-N resulted in virus yield similar to levels reported for deletion of the entire me53 (de Jong, 2011). In addition, deletion of the L region from aa 278-302 results in similar reduced BV production levels at 72 hpt (de Jong, 2011). These results suggest that the nuclear localization of ME53, its N-terminal zinc finger, and L region are essential for the nascent function of the polypeptide, and the C-terminal zinc finger, which results in a 2-fold decrease in BV production when deleted, is also potentially important in the production of BV. The decreased titres of deletion mutants highlight the importance of several conserved domains in ME53 function for overall BV production. However, the 24 hpt initial titre calculation does not examine the role of these domains during early times post infection, which could provide more insight into the roles of these domains in ME53’s function.

ME53 intracellular localization

ME53 is a minor component of the BV nucleocapsid and is localized in the cytoplasm upon infection and early gene expression. Nuclear translocation of ME53, observed by using a C-terminal green fluorescent protein (GFP) tag, begins at 6 hpi. Because its translocation is infection- and NTS-dependent it is likely dependent on an unidentified viral chaperone protein that binds to the NTS (Liu et al., 2016). In the nucleus, ME53

20 localizes to the ring zone, a less dense area surrounding the virogenic stroma where polyhedra begin to form, and it remains there until at least 36 hpi (de Jong et al., 2011;

Ozsahin, 2018 unpublished). During the late phase of infection, from 12 to 36 hpi, ME53 localizes to the plasma membrane and forms foci. Similar to nuclear translocation, foci formation is infection dependent. Specifically, the BV envelope protein GP64 is required for the localization of ME53 at presumed budding sites at the plasma membrane (de Jong et al., 2011).

Some work has been completed regarding the role of the aforementioned domains in intracellular localization of ME53. The NTS is essential for ME53 nuclear localization but it has not been determined if it is important for foci formation (Liu et al., 2016). The ZnF-

N may aid in the interaction with the putative viral chaperone protein that facilitates nuclear translocation, because deletion of a fragment of the ZnF-N (aa 160-191) reduces

ME53 nuclear translocation to 76% of WT (Liu, 2015). The reduction is not significant but suggests that the ZnF-N may serve a stabilizing structural role for the interaction of the

NTS with the chaperone or nuclear pore complexes. It is currently unknown if ZnF-N affects plasma membrane foci formation. Deletion of the ZnF-C domain does not affect nuclear localization or foci formation of ME53 indicating its putative function is not associated with the implicated viral chaperone proteins for either process (de Jong et al.,

2011). The intracellular localization of ME53 throughout infection, and the association of several conserved domains with varying effects on ME53 localization, supports the idea that ME53 may have distinct functions associated with each domain that are distinguished by the stage of viral infection and subcellular localization of ME53.

21

The association of ME53 with the capsid and its role in DNA replication

BV nucleocapsids assemble, without an envelope, in the nucleus then proceed to the plasma membrane where they gain their envelope from the plasma membrane through budding. ME53 is incorporated into the nucleocapsid of BVs during virus infection. Using Western blot analysis of purified virions and mass spectrometry, ME53 was detected in the BVs but results are contradictory regarding the association of ME53 with ODV nucleocapsids (de Jong et al., 2009; Wang et al., 2010; Hou et al., 2012).

AcMNPV ME53 was identified as a component of ODVs and the nucleocapsid of BVs via

Western blot analysis of purified virions. In comparison, using four separate spectrophotometric analysis tools, Group II alphabaculovirus Helicoverpa armigera nucleopolyhedrovirus (HearNPV) ME53 was found to be incorporated into only the nucleocapsid of BVs and not ODVs (Hou et al., 2012). The discrepancy is likely due to the different viruses and cell lines (AcMNPV versus HearNPV and Sf21 versus HzAMI, respectively), and the relative sensitivities of the two approaches. The assosciation of

ME53 with BVs supports the observation of foci formation and implicates ME53 in a DNA packaging or nucleocapsid assembly role in the nucleus during late times post infection.

As further support for the involvement of ME53 in virion packaging, ME53 does not affect viral DNA replication during infection, and its deletion results in an initial accumulation of viral DNA (de Jong et al., 2009). Deletion of me53 results in a 102-fold increase of intracellular DNA compared to a 16.5-fold increase for WT transfected cells

(de Jong et al., 2009). The smaller fold change of intracellular DNA in WT transfected cells is attributed to BV production and the corresponding loss of viral DNA into BVs. The significant accumulation of intracellular DNA indicates that ME53 is not required for DNA

22 replication but is likely involved in DNA egress via nucleocapsid assembly, nucleocapsid egress from the nucleus, nucleocapsid egress from the cell, or all of the above.

ME53 and the transcription of viral genes

It is surmised that ME53 is a transcriptional regulator for late genes due to its nuclear localization at late times post infection, and its two putative C4 zinc finger domains. To investigate the role that ME53 may play as a transcriptional regulator, Liu

(2015) chose 30 viral genes, grouped by function, to amplify and quantify using qRT-

PCR. The genes included: immediate early transcription factor genes (pe38, ie-1, and pp31/39k), genes essential for viral RNA polymerase complex formation (lef-4, lef-8, and lef-9), nucleocapsid structural proteins (p6.9, vp39, and p78/83), genes essential for nucleocapsid assembly (vp1054, 49k, 38k, odv-ec27, bv-odv,c42, ac109, vlf-1, and ac53), genes essential for nucleocapsid egress (exon0, and vp80), BV envelope proteins (gp64, odv-e18, ac124, pep, and ac73), ODV envelope proteins (odv-e56, bv-odv-e26, and odv- e25), and genes encoding non-essential proteins (chiA, v-cath, and iap2).

Based on ME53 knock-out studies the presence of ME53 had limited effect on the transcript levels of immediate early genes ie1, 39k, and pe38, that have known transcriptional regulatory activity (Liu, 2015). The viral RNA polymerase subunit genes were also unaffected by ME53 with transcript fold changes relative to wild-type for lef-4, lef-8, and lef-9 ranging between 1.2 to 2.0 at 18 and 24 hpt respectively. The transcription of viral genes essential for nucleocapsid assembly and nuclear egress increased up to 6- fold at 18 and 24 hpt in the presence of ME53 (WT). The specific genes include p78/83, ac53, vlf-1, vp39, and vp80. The capsid basal structural protein p78/83 transcript levels were down-regulated by 2.5-fold and 3.7-fold at 18 and 24 hpt respectively. In the

23 absence of me53, the nucleocapsid assembly genes ac53 and vlf-1 were affected the most. The presence of ME53 resulted in a 4.7-fold increase of ac53 transcripts at 18 hpt and a 6.0-fold increase at 24 hpt. Vlf-1 transcripts showed a 4.1-fold increase at 18 hpt, and a 2.7-fold increase at 24 hpt. The major capsid protein vp39 transcripts increased by

3.7-fold and 2.4-fold at 18 and 24 hpt respectively, and vp80, an essential gene for viral egress from the nucleus, increased 3.8-fold and 3.9-fold at 18 and 24 hpt (Liu, 2015). The influence that ME53 has on late genes may be due to several mechanisms (Liu, 2015).

ME53 may directly up-regulate genes involved in the transactivation of other genes or genes associated with the viral RNA polymerase complex and cause indirect upregulation of late genes. Secondly, although ME53 does up-regulate other viral immediate early and late genes, it may directly up-regulate viral late gene expression (Liu, 2015). There may also be an interaction of ME53 up-regulating late genes directly and indirectly through the early genes (indirectly) and the late genes (directly) (Liu, 2015). Considering that early genes were up regulated only ~2-fold and late genes were ~4-fold higher in the presence of ME53, ME53 might have both direct and indirect influences on late gene expression.

C4 zinc finger proteins

Zinc finger motifs are found in numerous DNA binding proteins and transcriptional regulators; however, they are also known to facilitate protein-protein and protein-RNA interactions (Pace and Weerapana, 2014). Most annotated zinc finger proteins have two cysteine and two histidine residues (C2H2) bound to zinc in a tetrahedral conformation such as in transcription factor IIIA (Figure 1.6) (Iuchi and Kuldell, 2005). Other combinations of residues include three cysteine residues and one histidine residue (C3H) or, as in the case of ME53, four cysteine residues (C4) (Iuchi and Kuldell, 2005; de Jong

24 et al., 2009). Zinc finger complexes can participate in a variety of functional roles including structural, catalytic, and regulatory roles, but are predominantly classified as transcriptional activators and repressors (Pace and Weerapana 2014). C4 zinc fingers are less likely to be involved in catalysis or enzymatic transformations because of the large charge transfer facilitated by the thiol group of cysteine rendering the Zn2+-cysteine complexes relatively static (Pace and Weerapana, 2014). Further classification of zinc finger domains by structure is accomplished using theoretical approaches, like homology modeling, and physical methods such as nuclear magnetic resonance or X-ray crystallography. Zinc finger function can also be suggested through structural homology, but the growing body of knowledge has demonstrated that zinc finger domains with similar structure can have several different functions. For example, the treble clef finger family is subdivided into ten groups with roles including phosphatidylinositol-3-phosphate binding, nucleic acid binding, and catalytic activity (Krishna et al., 2003). ME53 contains two putative C4 zinc finger domains separated by 170 amino acids. Proteins with multiple zinc finger domains that participate in the same function are typically separated by less than

50 aa (Krishna et al., 2003). For example, OAZ is a 30 aa-zinc finger protein that uses a distinct group of zinc fingers to bind DNA, shown in Figure 1.6, and the other zinc finger clusters for different protein-protein interactions (Hata et al., 2000). The smallest number of amino acid residues between the functionally distinct groups of zinc fingers is 57 aa, while the average number of amino acid residues between the zinc fingers of OAZ involved in the same function is 12 (Hata et al., 2000). The 170 amino acids separating the putative C4 zinc fingers of ME53 suggests two functionally distinct roles. However, the

25

Figure 1.6: Transcription factor IIIA from Xenopus laevis The first-identified zinc finger domain was in transcription factor IIIA from Xenopus laevis oocytes (Miller et al., 1985). The region in the red box shows one zinc finger in cyan bound to the major groove of the double-stranded DNA in orange. The amino acids shown as sticks represent the two cysteine residues, with the sulfur atoms highlighted in yellow, and the two histidine residues with the nitrogen atoms shown in blue that coordinate a zinc ion. The zinc ion, shown as the grey sphere is coordinated by the atoms on the cysteine and histidine residues (PDB ID 1TF6).

26 tertiary structure of the protein could bring them within contact to be functionally cooperative.

A family of nuclear hormone receptors is characterized by the presence of two C4 zinc finger domains that have discrete functions as receptors in the cytoplasm or nucleus

(Zhang et al., 2004). The cytoplasmic receptors bind DNA sequences in response to ligand binding. Functionally, a highly conserved DNA-binding domain is joined by a linker sequence connecting a C-terminal ligand-binding domain (Zhang et al., 2004). The DNA binding domain consists of two zinc-finger domains with sequence-specific binding affinity controlled by the amino terminal zinc finger and the carboxyl terminal zinc finger modulating dimerization (Zhang et al., 2004). Preliminary yeast-2-hybrid data suggests that ME53 does not interact with itself and therefore would not dimerize. Currently, no homologues of ME53 have been identified therefore hindering determinations of homology-modeling software. The confirmation of zinc finger domains in ME53 and their structure would be an asset in elucidating the function of these domains during baculovirus infection.

Thesis Objectives

The AcMNPV protein ME53 is required for optimal budded virus production in cell culture. Although several studies have determined its significance in overall virus production and tracked its subcellular localization, questions still remain regarding its function during virus infection. The highly conserved regions in ME53 including two putative C4 zinc finger domains with 100% conserved cysteine residues are likely to be functionally significant; however, the ability of these regions to actually bind zinc remains

27

to be determined. The first objective of this thesis is to determine if the putative C4 zinc fingers coordinate zinc ions.

The role of ME53 in baculovirus infection over several cycles of virus replication has been determined, but the role of ME53 and its conserved domains, putative zinc fingers, and associated cysteines during the initial (not secondary) viral infection cycle

<24 hpi have not. The initial viral infection cycle can suggest the relative location of ME53 domain function as the protein translocates to different subcellular regions during infection. For example, if deletion of a domain affects virus production after 18 hpi, it suggests that the domain in question is important in the nucleus of infected cells, or for foci formation or budding, as ME53 is localized in the nucleus and plasma membrane foci at this time point. Zinc finger domains are traditionally thought to be associated with transcription factors, and although deletion of the entire ME53 has minimal effects on virus gene transcription, the effect on viral gene transcription by deletion of the putative zinc finger domains or linker has not been determined. The second objective of this thesis is to elucidate the role of several conserved domains of ME53 and the conserved cysteine residues of the putative zinc fingers in the early stages of virus production, and their role in virus gene transcription at early and late times.

Currently, the mechanism by which the host anti-viral defense of global translation shutdown is regulated by baculovirus infection is unknown. The association of anti- apoptotic proteins with translational inhibition suggests that similar events such as viral

DNA replication and host gene shut off trigger the defense pathway. Several immediate early proteins instigate apoptosis when transiently expressed and indicate that an immediate early protein may be involved in translational regulation. Immediate early

28 protein ME53 interacts with the ribosomal protein RACK1, which is frequently targeted for sequestration during proteotoxic stress thereby inhibiting translation. The nature of the

ME53-RACK1 interaction is currently unknown so the third and final objective of this thesis is to determine if ME53 associates with host translational machinery during infection and to begin to characterize its interaction with RACK1.

29

Chapter 2: Bioinformatic analysis of ME53 secondary structure and confirmation of zinc finger domains, and ME53 overexpression

Introduction

Zinc is the second most abundant transition metal in biological systems after iron and unlike other first-row transition metals, has a filled d-10 orbital (Cao et al., 2017). Zinc exclusively forms a Zn2+ ion that typically lacks redox activity and is an important ligand for the structural stability of proteins with zinc finger motifs representing the most common structural domain associated with zinc (Cao et al., 2017). Although Zn2+ is considered relatively static, several enzymes use a Zn2+ ion for enzymatic purposes with hydrolysis and condensation reactions being the most frequent (Laitaoja et al., 2013). This is due to the flexibility of Zn2+ coordination and its net charge of +2 generating deprotonated binding partners, such as water and the cysteine side chain, that act as nucleophiles in catalytic reactions (Cox and McLendon, 2000).

Zinc finger proteins are some of the most abundant proteins in eukaryotes with an estimated 2% of all proteins containing a zinc binding motif of conserved cysteine and histidine residues. A zinc ion is typically coordinated by either of two cysteine and two histidine residues (C2H2), three cysteine residues and one histidine residue (C3H), or four cysteine residues (C4) (Perales-Calvo et al., 2015). Proteins containing these domains can have several zinc fingers in close proximity, such as the DNA-binding domains of transcription factor IIIA (TFIIIA), or they can be the only domain present with the average motif encompassing 20-30 amino acids that protrude as a finger-like structure around the

Zn2+ ligand (Grishin, 2001). Zinc fingers tend to be synonymous with DNA-binding transcription factors, but are associated with RNA-binding, protein-protein interactions, regulation of apoptosis, membrane association, and protein folding (Lebrun et al., 2016).

30

The majority of zinc finger families are classified by the ligand-coordinating residues, such as C2H2, and the secondary structure they assume, and are named for the first gene/protein identified to have that structure; however, the naming conventions result in ambiguous classifications. Substantial work to classify zinc finger domains by their secondary structure resulted in 8 different fold groups (Krishna, 2003). The first fold identified in TFIIIA is the C2H2-like fold that is found in C2H2, C3H, and IAP domains where a -hairpin is followed by an -helix (Krishna, 2003). The gag knuckle fold group includes the retrovirus zinc finger-like domains where two short -strands are connected by a turn and followed by a short -helix or loop (Krishna, 2003). The largest fold group is the treble clef group where a -hairpin at the N-terminus and an -helix at the C-terminus are separated by a loop or a -hairpin, resulting in a ribbon structure resembling a treble clef.

The treble clef group includes RING finger domains, such as those found in AcMNPV immediate early protein IE2 and PE38 (Prikhod’ko and Miller, 1998). The second largest fold group is the zinc ribbon fold found in the Adenovirus early E2A DNA-binding protein, and ribosomal protein L36 (Krishna, 2003). The remaining groups are the Zn2/Cys6 fold found in the GAL4 transcription factor, Taz2 domain found in the N-terminal zinc binding region of HIV integrase, and the less common zinc binding loops and metallothioneins that have no regular secondary structure but a similar spacing of the zinc binding residues.

The predominant methods currently used to confirm zinc fingers are structural determination of the regions using X-ray crystallography or nuclear magnetic resonance imaging. However, zinc binding by these regions can be inferred by their structural dependence on zinc and through sequence and structural dependence, zinc finger

31 domains can be confidently predicted. Circular dichroism (CD) spectroscopy is a simple method to analyze the overall secondary structure of a protein or peptide under varying conditions. In the case of zinc finger domains, CD can be used to observe the change in residue angles in the presence and absence of zinc due to the effect of zinc on the structural conformation of these domains via the zinc binding residues (Kelly et al., 2005;

Greenfield, 2006). The presence of conserved cysteine residues with spacing appropriate to zinc fingers (C-X2-C-X15-30-C-X2-C) in two regions of ME53 suggests that these regions are capable of coordinating a zinc ion. In this chapter, peptides of these regions, as opposed to whole protein, were used to increase the sensitivity of the CD analysis and to observe the influence of zinc on these regions to confidently predict the presence or absence of two zinc finger domains in ME53.

Methods

2.2.1 Bioinformatic analysis of ME53 disorder, binding prediction, and redox potential

For bioinformatic analysis of ME53 structure, the amino acid sequence used is that from accession number AAA46718.1. Initial Blast searches for homologues of whole

ME53, as well as peptides covering residues 170-209 and 379-399, returned no reliable

(E-value < 0.01) homologues for the inference of structural data. Instead, IUPred2A

(https://iupred2a.elte.hu) was initially used to predict the global structural disorder of

ME53. The structure of domains disorder prediction algorithm was used to identify disordered regions of ME53 as well as continuous ordered regions of at least 30 aa. The

ANCHOR2 algorithm was applied to predict the probability of disordered binding domains.

The probability values from the ANCHOR2 output were also used to infer the likelihood of binding sites in ordered regions of ME53 due to the algorithm’s prediction of the NTS

32

(residues 109-137) to act as a binding domain. There is experimental evidence that the

NTS is a binding domain for a chaperone protein for nuclear translocation and so, the

ANCHOR2 probability values for the NTS were considered acceptable for other ordered domains. IUPred2A also identifies potential redox sensitive regions that can be ordered or disordered depending on the redox state for the environment. The output scores for redox potential were compared against the IUPred2A scores and a redox potential value higher than the IUPred2A value was considered predictive of a redox-sensitive region.

The secondary structure prediction server JPred4

(http://www.compbio.dundee.ac.uk/jpred4/help.shtml) was used following IUPred2A. The same input sequence used for IUPred2A analysis of ME53 was also used for JPred4. The predictions included in the output were jnetpred secondary structure cartoon, the

JNETCONF scores to show the confidence values associated with the secondary structure prediction, and the Jnet Burial histogram that represents the probability of residues to be buried.

2.2.2 Cell lines

Sf9 cells, derived from the fall army worm Spodoptera frugiperda, were maintained as monolayer cultures in tissue culture flasks at 27°C in Hink’s TNM-FH medium (Wisent

Inc.) (van Oers, 2010). Cell culture medium was supplemented with 10% fetal bovine serum (FBS) (Invitrogen) unless otherwise stated.

2.2.3 Peptide resuspension

Peptides used for CD, listed in Table 2.1 were synthesized by Applied Biological

Materials and confirmed to be greater than 95% pure by high-performance liquid chromatography. Lyophilized peptides were put in solution at 0.5 mg/mL and 1 mg/mL,

33

Table 2.1 ME53 Zinc Finger Peptides Corresponding Peptide AcMNPV ME53 Amino Peptide Sequence Acids N-SRCTTCNYRFKDNTREWFLYVWHIEKPLDD ZnF-N 168 – 211 PDRIDICCQKCYL-C ZnF-C 377 – 401 N-NYCKLCKKTKLYYKNPVLYCTKCGF-C

34 for the ZnF-N peptide and ZnF-C peptide respectively, in 20 mM Tris (pH 7.8) and 0.1 mM Tris (2-carboxyethyl) phosphine (TCEP) to prevent oxidation of the cysteines prior to incubation with ZnCl2.

2.2.4 Circular dichroism

The CD spectra were acquired for the ZnF-N and ZnF-C peptides on a JASCO J-

815 CD spectropolarimeter using a 1-mm path length quartz CD cuvette. The peptides were at a final concentration of 0.16 mg/mL in 20 mM Tris (pH 7.8) and were incubated in a final concentration of 150 M ZnCl2 in 20 mM Tris for 10 minutes at room temperature priorto the addition of 0.1 mM EDTA as indicated. The temperature of the sample cells was controlled by a circulating water unit set to 21ºC and each scan was corrected against buffer-only spectra to account for signal due to the buffer. Each spectrum is the average of eight scans from 190–250 nm. The data was averaged and smoothed using the

Savitzky-Golay algorithm in the Jasco data analysis package.

2.2.5 Analysis of CD data

The CD Analysis and Plotting Tool (CAPITO) online platform

(http://capito.nmr.leibniz-fli.de) was used for the conversion of CD data from millidegrees

(mdeg) to molar ellipticity for the comparison of ME53 peptide spectra with CD spectra of other proteins and peptides of known structure. CAPITO uses the CD spectra of 107 proteins with associated Protein Data Bank (PDB) files for secondary structure prediction, and 95 datasets of CD values at 200 nm and 222 nm to infer unfolded or folded states, respectively. The secondary structures included in the analysis are -helix, which includes -helix, 310-helix, and π-helix, and -strand, which includes parallel and antiparallel strands and -bridges, and irregular, which includes bonded turns, bends,

35 loops, and irregular structures. CAPITO reliably predicts -helical and -strand secondary structures as well as irregular structures, albeit slightly less accurately, and can therefore be used to observe the change in secondary structure of the ME53 ZnF-N and ZnF-C peptides under changing conditions.

2.2.6 Construction of ME53 overexpression bacmid

The baculovirus expression system was used to overexpress His-tagged ME53 in

Sf9 cells for future use as a reagent for binding studies. The plasmid pFastBacHTB

(Invitrogen) was used as the donor plasmid because of its Tn7 cassette that contains a multiple cloning site downstream of several features including: the baculovirus polyhedrin promoter, a 6x histidine (His) tag and a tobacco etch virus protease cleavage site to remove the His tag following purification. Me53 was amplified from the bMON14272 bacmid genome using forward primer me53-FW 5’-CGGGATCCAACCGTTTTTTTC -3’

(the BamHI site is underlined) and reverse primer me53-RV 5’-

CGGAATTCTTAGACATTGTTATTTACAATATTAATTAAC-3’ (the EcoRI site is underlined). The insert and vector were cleaved with EcoRI and BamHI in 1X Fast Digest

Buffer (Thermo Fisher) for 20 min at 4°C and ligated using T4 Ligase (Thermo Fisher) in

1X T4 Ligase buffer (Thermo Fisher) at 4°C overnight. Following transformation into E. coli DH5 cells, the plasmid DNA was isolated and sequenced to confirm the successful insertion of ME53 and to ensure that no mutation, such as introduction of a stop codon, arose during amplification. The sequenced plasmid was subsequently transformed into chemically competent E. coli DH10 cells that contained the bMON14272 bacmid as well as a helper plasmid encoding a transposase for transposition of the pFastBacB:ME53

36

Tn7 cassette into the atti-Tn7 site of the bacmid. His-ME53 insertion and expression was confirmed via PCR and Western blot respectively.

2.2.7 Bacmid DNA isolation

Recombinant bacmid DNA was isolated from transformed E. coli DH10 cells as described in the Bac-to-Bac Baculovirus Expression Systems Manual (Invitrogen). Cells were pelleted by centrifugation at 14,000 x g for 1 minute. The cell pellet was resuspended in 300 L 15 mM Tris-Cl, 10 mM EDTA and 100 g/mL of RNase A. Cells were lysed for

5 minutes at room temperature by adding 300 L of 0.2 N NaOH and 1% SDS. Three hundred L of a 3 M potassium acetate, pH 5.5, solution was added to precipitate proteins, and samples were incubated on ice for 10 minutes. The samples were centrifuged at 14,000 x g for 10 minutes to pellet any protein. The supernatant was transferred to a new microcentrifuge tube containing 800 L of absolute isopropanol to precipitate bacmid DNA, and incubated on ice for 10 minutes. Then, the samples were centrifuged at 14,000 x g for 15 minutes to pellet the DNA precipitate. Supernatant was removed, and the pellet was washed with 500 L of 70% ethanol and centrifuged at

14,000 x g for 5 minutes. Supernatant was removed by pipette from the pellet which was then air-dried before being resuspended in 40 L of Milli Q H2O (mqH2O) and quantified using a NanoDrop ND-1000 (Thermo Fisher).

2.2.8 Virus amplification

The recombinant Ac+HisME53 virus was amplified by first transfecting 1.0 x 106 cells seeded in 35 mm dishes with 5 g of purified bacmid DNA using 8 L of Cellfectin II reagent. The bacmid DNA and Cellfectin II were each independently diluted into 100 L of Grace’s Serum Free medium (SFM) before being combined and incubated at room

37 temperature for 30 minutes. The Sf9 cells were rinsed with 1 mL of Grace’s SFM prior to the addition of the bacmid DNA/Cellfectin II solution in 0.8 mLGrace’s SFM. The cells and

DNA/Cellfectin II solution were incubated at 27°C for 5 hours before the solution was removed and replaced with fresh Hink’s medium +FBS. The supernatant of transfected cells was removed at 72 hpi and kept as the P0 virus stock. The P0 stock was used for virus amplification by seeding 1.0 x 107 Sf9 cells in a T75 flask allowing the cells to adhere for 1 hour and replacing the medium with 200 L of the P0 stock diluted in two mLof Hink’s

+ FBS medium. The cells were left on a shaker for two hours before 15 mL of fresh Hink’s

+FBS medium was added. After 72 hours, the supernatant from infected cells in the T75 flasks was removed and centrifuged at 1,000 x g for 5 minutes and the cleared supernatant was transferred to a new 15 mL tube and stored at 4°C in the dark. This BV- containing supernatant is referred to as the P1 virus stock and was used for subsequent infections.

2.2.9 Estimation of viral titre using the end-point dilution assay

Viral titre was determined using the TCID50 end-point dilution method (Reed and

Muench, 1983). Briefly, virus supernatant was serially diluted from 100-10-9 and 100 L of each dilution was added to each of eight wells of a 96-well plate seeded with 100 L per well of Sf9 cells at a density of 1.0 x 105 cells/ml. Cells in eight wells of the infected plates were left uninfected as a negative control. The 96-well plates were incubated at

27°C for 7 days before they were scored for the presence of virus based on cytopathic effects (CPE) that include cellular debris from cells that have undergone apoptosis, enlarged nuclei, membrane blebbing, and limited to no growth of cells (Rohrmann, 2013).

Titre was calculated by the Reed-Muench method where the tissue culture infectious dose

38 causing a 50% rate of infection was used for titer measurement and reported as

TCID50/mL.

2.2.10 Multiplicity of infection calculation

A given multiplicity of infection (MOI) was used to infect cells with a consistent number of infectious viral particles per cell. The equation used was modified from the

Bac-to-Bac expression manual (Invitrogen):

Inoculum required (mL) = desired MOI (virus particles/cell) x (Total number of cells) Titre of viral inoculum (TCID50/mL)

The desired MOI represents the number of virus particles per cell that the cell culture will be infected with and determines the necessary volume of the virus stock needed to infect a number of cells at the chosen MOI.

2.2.11 His-ME53 purification

Sf9 cells were seeded to a density of 1.0 x 107 cells/mLin T75 flasks, allowed to attach for one hour, and infected with Ac+HisME53 at an MOI of 10. Medium was removed from the cells and replaced with the appropriate volume of virus solution. Flasks were left to gently rock for 2 hours before the virus solution was removed and replaced with 15 mLof Hink’s +FBS medium. At 48 hpi the cells and supernatant were collected and pelleted at 1,000 x g for 5 minutes. The supernatant was decanted, and the cells were washed in ice-cold phosphate buffered saline (PBS) and recentrifuged. Cells were lysed in 4 mLlysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, pH 8.0) and supplemented with 1% Igepal (NP-40) (Sigma-Aldrich), which keeps nuclei intact. This was to prevent contamination of the sample with nuclear proteins. PMSF (Thermo

Scientific) at 0.1% was added to the lysate, which remained on ice for 10 minutes. The lysate was centrifuged at 10,000 x g at 4°C for 10 mins to pellet the nuclei and membrane

39 components. Two hundred L of a Ni-NTA (Qiagen) 50% slurry was equilibrated with lysis buffer then mixed with the supernatant and incubated on a nutator (Thermo Fisher) at

4°C for 2 hours. The lysate and Ni-NTA mixture was loaded into a 4.5 cm polypropylene column (BioRad) and the flow through was collected. Eight hundred L of wash buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole) was passed twice through the column and collected followed by 100 L of elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole) to elute His-tagged ME53.

2.2.12 SDS-PAGE and Western blot analysis

Eluates from His-ME53 purification were separated in 12% SDS-PAGE gels for 2 h at 100V before being Coomassie blue stained or transferred to polyvinylidene difluoride

(PVDF) membrane for 1 h at 30 mA in fresh buffer (~100V). The PageRuler pre-stained protein ladder (Thermo Fisher) was used to determine the molecular weight of proteins in

Coomassie blue or Western immunoblot bands. For Coomassie blue staining, gels were stained with 0.2% Coomassie Blue R-250 (Fisher) in 10% acetic acid, 50% methanol, and

40% dH2O for 30 minutes. The gels were destained in 50°C dH2O for 15 minutes 4 times with shaking. For Western blotting, following protein transfer to the PVDF membranes, the membranes were incubated in blocking buffer consisting of 2% skim milk powder in tris-buffered saline (TBS) overnight at 4°C. After blocking, the membranes were rinsed in

TBS with 0.1% Tween-20 (TBS-T) for 10 minutes three times. The membranes were then incubated with primary monoclonal mouse anti-His antibody (Sigma, H1029) for

His:ME53 detection at a 1:3,000 dilution in TBS-T with 0.2% skim milk powder. Primary antibody incubation with the membrane was carried out at 20°C with gentle shaking for 1 hour. The membrane was then washed in TBS-T for 5 mins 4 times before being

40 incubated with the horseradish peroxidase (HRP)-conjugated secondary antibody, rabbit anti-mouse IgG (Invitrogen) in TBS-T with 0.2% skim milk powder. Secondary antibody incubation occurred over 1 h at room temperature with shaking and was followed by washing the blot with TBS-T for 10 min 4 times. Super signal West Pico chemiluminescent substrate (Thermo Fisher) was applied to the membrane to detect the HRP-conjugated secondary antibodies for the Western blot. The ChemiDoc MP System (BioRad,

Mississauga, ON) was used to detect the fluorescence and visualize the bands.

Results

2.3.1 Structural Analysis of ME53

Secondary structure and intrinsic disorder prediction softwares, IUPred2A, and

JPred respectively, were used to infer the order and secondary structure of AcMNPV

ME53. IUPred2A uses energy estimations of each amino acid in a sequence to quantify their interactions within the protein as favourable or unfavourable (Meszaros et al., 2018).

Unfavourable residue energy is associated with disorder, while favourable energy is characterized as ordered. The energy values are then output as a value between 0 and

1 with the higher values associated with increased probabilities of disorder, and values above 0.5 meaning that the residue is disordered. IUPred2A also uses the ANCHOR2 algorithm to predict available binding sites in disordered regions on the same scale as the order/disorder output (Figure 2.1A). A second option in IUPred2A shows potential redox sensitive regions using an algorithm that swaps serine residues for cysteine residues and observes the change in order/disorder (Figure 2.1B).

The IUPred2A provided output indicated a high disorder prospensity for the first 83 amino acids of the AcMNPV ME53 N-terminus (Figure 2.1A). The remaining regions all

41

NTS ZnF-N L ZnF-C A

B

Figure 2.1: IUPred2A analysis of ME53. Disorder, binding, and redox prediction from IUPredA for ME53 (schematic shown above). (A) ME53 disorder analysis with the probability of disorder shown in red (IUPred2A) and the probability of a disordered binding site shown in blue (ANCHOR2). (B) Prediction of redox potential, shown in purple, compared to the IUPred2 prediction of disorder in red.

42 had a low probability of disorder and a general globular structure for ME53 from residues

84-449 was predicted. The highly conserved regions of ME53, NTS (aa 109-137), ZnF-N

(aa 170-209), L (aa 278-302), and ZnF-C (aa 379-399), were located within the predicted globular structure of ME53. The NTS disorder probability values ranged from 0.2-0.4 suggesting a relatively ordered structure with a slightly more ordered region towards the

C-terminus (aa 126-137) of the domain. The putative ZnF-N motif was unlikely to be disordered with values ranging from 0.02-0.19 and 66% of the residues had disorder probability values below 0.1. The conserved L region was also relatively ordered with low disorder probability values from 0.1-0.26. Finally, the putative ZnF-C domain had disorder probability values similar to the ZnF-N motif and was confidently predicted to be ordered.

The ANCHOR2 algorithm output showing the predicted disordered binding regions is overlayed with the IUPred2A output (Figure 2.1A). Since ME53 was predicted as mostly ordered, the ANCHOR2 output values were lower than the default 0.5 threshold.

However, a commonality between ordered and disordered binding regions is the conformational change that binding regions typically undergo upon ligand binding.

Therefore, when the ANCHOR2 values for ME53 positively diverged from the IUPred2A values, a putative binding site could be inferred. The NTS domain of ME53 is a presumed binding domain for a host or viral chaperone protein that facilitates ME53 nuclear translocation. This domain is required for the translocation of ME53 to the nucleus and was predicted as a binding domain by ANCHOR2 as the probability values increased over the IUPred2A disorder prospensity values. Other domains of ME53 that showed similar potential were residues 184-194 within the putative ZnF-N, and a large region from

43 residues 293-341 that includes the C-terminal half of the L region. The ZnF-C showed no potential binding regions, with ANCHOR2 values close to 0.

The prediction of redox potential for ME53 showed potential redox sensitive regions at the cysteine residues for each putative zinc finger domain (aa 170-209, 379-

399) (Figure 2.1B). The putative binding region from residues 293-341 was also predicted to have redox potential likely due to the presence of cysteine residues at positions 314 and 325. ME53 contains twelve cysteine residues, eight of which are associated with the putative C4 zinc finger domains. The cysteine residues not associated with the putative zinc fingers are located at residues 205, 265, 314, and 325. Two potential redox sensitive regions correlated with putative binding regions, the ZnF-N, and the large region C- terminal to the L domain from amino acids 293-341.

ME53 was predicted to be largely ordered so JPred was used to predict its secondary structure (Figure 2.2). The region of disorder in the first 100 amino acids of the N-terminus predicted by IUPred2A was also predicted with relatively high confidence to be a coiled region by JPred. The remainder of ME53 was predicted to be composed of predominantly -strands with the identified NTS predicted to be a single -helix. The putative ZnF-N was predicted to have two -strands but only the central strand (aa 184-

191) was associated with high confidence values as well as the predicted binding region.

The remainder of the ZnF-N was predicted to be a coil structure and correlates with solvent accessible residues. The putative ZnF-C was predicted to be comprised of three

-strands however, only two -strands encompassing the N-terminal and C-terminal cysteine pairs had confidence values over 5. The remaining residues of the ZnF-C were

44

Figure 2.2 JNet Pred Output for ME53. JPred4 output for ME53 showing the amino acid sequence and schematic of domains above. The graphic beneath the sequence labeled Jnetpred shows the predicted secondary structure with the grey line indicating coil, red indicating -helix, and -sheet depicted by the green arrows. The JNETCONF histogram indicates the confidence of the algorithm in the secondary structure prediction with 9 representing high confidence in the prediction. The Jnet Burial histogram represents the probability of residues being buried. The larger the bar, the less accessible the residue is to solvent.

45

Table 2.2: Summary of predictions from the IUPred2A and JPred4 servers. The cartoon schematic for JPred secondary structure (ss) displays either alpha helices (red) or beta strands (green arrow) with the asterisk below indicating >50% confidence in the prediction.

Redox Domain Residues IUPred2A ANCHOR2 JPred4 ss Potential Binding site NTS 109-137 Ordered None (aa 122-139) * Binding site ZnF-N 170-209 Ordered Yes (aa 184-194) * Binding site L 278-302 Ordered Some (aa 293-341) * * No binding ZnF-C 379-399 Ordered Yes site * *

46 confidently predicted to be coiled. A summary of the algorithm predictions is shown in

Table 2.2.

2.3.2 Circular dichroism analysis of ME53 peptides

Circular dichroism spectroscopy was used to qualitatively determine the structure of ME53 peptides encoding the putative C4 zinc finger domains in the presence and absence of Zn2+ ions. CD spectra have characteristic minima and maxima depending on the secondary structure of the protein or peptide being measured (Greenfield, 2006).

Disordered proteins have elliptical minima near 198 nm and ellipticity values close to zero at 222 nm. Proteins that are composed predominantly of -sheets show a minimum at

210-218 nm and a maximum at 195 nm, whereas globular proteins with a dominant - helix structure show minima at 222 nm and 208 nm, with a large positive maximum at 195 nm respectively. The online CD Analysis and Plotting Tool (CAPITO) uses CD reference banks and the difference in area between known and sample spectra to estimate the secondary structure of a protein.

The JPred output estimated that the ZnF-N region within the whole protein would be composed of 32% -strand. The CD spectra for the ZnF-N peptide (aa 168-211) in buffer had a strong minimum at 195 nm, characteristic of a disordered protein (Figure

2.3). The addition of 150 M of ZnCl2, a 5.06-fold molar ratio of zinc to peptide, resulted in a CD spectrum with a minimum at 212 nm, consistent with a -sheet structure. CAPITO analysis of the peptide spectra without zinc predicted a mostly disordered peptide with

31% -strand structure (NRMSD 0.20). The addition of zinc increased the -strand prediction to 36% (NRMSD 0.32) indicating that zinc increased the amount of secondary structure in the peptide. The CD spectra of the ZnF-N peptide shifted from disordered to

47 an apparent -strand conformation with the addition of zinc demonstrating that zinc influenced the structure of the peptide, presumably by forming a zinc finger.

The solution CD spectra for the ZnF-C peptide (aa 377-401) in the absence of zinc showed a characteristic disordered spectrum with a negative minimum at 195 nm (Figure

2.4). The addition of 150 M ZnCl2, a 2.83-fold molar ratio, caused a shift in the spectrum to resemble that of a globular -helix protein with negative minima at 203 nm and 222 nm, respectively. This is contradictory to the JPred prediction, which indicated that the

ZnF-C region was composed of 55% -strand. In the absence of ZnCl2, CAPITO estimated that only 31% (NRMSD 0.14) of the peptide contained -strand structure. The addition of ZnCl2 changed the -strand composition to an estimated 37% (NRMSD 0.77) indicating a gain in structure with ZnCl2. However, the NRMSD value for the peptide +

ZnCl2 spectra was high, so there is limited confidence in the prediction. CAPITO returned ubiquitin as the most similar spectrum based on the frequency of return and the side by side comparsion of the ME53 ZnF-C peptide spectra and Ubiquitin (Figure 2.5). Ubiquitin is considered a “not so straightforward case” by CAPITO as it is difficult to estimate its secondary structure from CD spectra. CAPITO estimates ubiquitin secondary structure as 11% -helix, and 32% -strand however its sequence suggests a 63% -helix and

32% -strand, which agrees with its crystal structure (Kumar et al., 1987). The discrepancy seen between the Chou-Fasman algorithm and the CD data interpretation of secondary structure for the 25 aa peptide encompassing the ZnF-C region is similar to the discrepancy seen with ubiquitin. The ZnF-C peptide data is directly compared to the

Ubiquitin data to show that secondary structure of peptides and small proteins is not always accurate. Although the secondary structure prediction between sequence and CD

48

Figure 2.3 Solution CD spectra analysis by CAPITO of the ZnF-N peptide in the presence and absence of zinc. (A) The CD spectrum for the ZnF-N peptide without zinc is shown in black, while the spectrum for the peptide with 150 M of ZnCl2 is shown in green. The spectra are shown in values of molar ellipticity from 190 – 250 nanometers. (B) The molar ellipticies of the peptides at 222 nm (y axis) and 200 nm (x axis) were used and compared against a database of CD spectra from proteins with known structure (PCDDB entries as grey filled squares), and a database of CD spectra from proteins with known disordered structure shown in clear boxes (Uversky et al. 2002). The black point shows the expected structure of the ZnF-N peptide in the absence of zinc, while the green point shows the estimated structure of the peptide in the presence of ZnCl2.

47

Figure 2.4 CD spectra analysis by CAPITO of the ZnF-C peptide in 20 mM Tris, pH 7.2, in the presence and absence of zinc. (A) The CD spectrum for the ZnF-C peptide without zinc is shown in black, while the spectrum for the peptide with a 3-fold molar excess of ZnCl2 (150 M) is shown in green. The spectra are shown in values of molar elipticity from 190 250 nanometers. (B) The molar ellipticies of the peptides at 222 nm on the y axis and 200 nm on the x axis are used and compared against a database of CD spectra from proteins with known structure (PCDDB entries shown in grey boxes) and a database of CD spectra from proteins with known disordered structure shown in clear boxes (Uversky et al. 2002). The black point shows the expected structure of the ZnF-C peptide in the absence of zinc, while the green point shows the estimated structure of the peptide in the presence of ZnCl2.

48 spectra are different, a qualitative shift from disordered to ordered was seen with the ZnF-

2+ C peptide in the presence of ZnCl2 indicating that Zn contributed to peptide structure.

2.3.3 Construction of His-ME53 overexpression bacmid

Further structural analysis of the ME53 zinc-coordinating domains requires full length ME53 protein. Peptides of the zinc finger domains were initially chosen, as opposed to designing an expression and purification system for ME53, to ensure that the structure of the zinc coordinating domains is influenced by the presence of zinc. With large proteins, like ME53, small structural changes due to, for example binding of zinc, may not significantly influence the overall secondary structure of the protein, making it challenging for CD analysis. Since the peptides’ structures seem to be dependent on zinc, an expression system for ME53 purification was designed and validated.

Whole protein structural analysis generally requires a substantial amount of purified protein from an overexpression system. Bacterial expression systems are often used for their high levels of expression and low input cost but can result in proteins lacking glycosylation and other post-translational modifications

(PTMs). The Baculovirus expression vector system (BEVS) is a eukaryotic system allowing for PTMs and exploits the viral polyhedrin promoter for overexpression of foreign mRNAs. The insect system was also chosen for overexpression of ME53 as that is its native environment.

The me53 ORF was successfully cloned into the donor plasmid pFastBac HT B downstream of the highly active polyhedrin promoter and a 6x His tag to facilitate purification of His-tagged ME53. The plasmid was then used to transpose me53 into the commercial bacmid genome (bMON14272) (Figure 2.6). Expression of HisME53 was

49

ZnF-C+ZnCl2 Secondary structure analysis

Predicted CD interpretation

ZnF-C +ZnCl2 a -0% b-68% i-32% a-13% b-38% i-58% Ubiquitin a-63% b-32% i-5% a-11% b-30% i-52%

Ubiquitin

Figure 2.5: Comparison of the secondary structure prediction similarities of AcMNPV ME53 ZnF-C and Ubiquitin CD spectra of the ZnF-C peptide in the presence of zinc (above), and Ubiquitin (below). The secondary structure predictions from the sequences using the Chou-Fasman algorithm and the CD data are shown in the table, respectively. The NMR structure of ubiquitin (PDB ID:1UBQ) is shown as a cartoon.

50 confirmed by transfecting the bacmid DNA into 1.0 x 106 Sf9 cells and performing a

Western blot with an anti-His antibody on the blot of the crude protein extract. Bacmids positive for HisME53 expression were then used to amplify Ac+HisME53 virus to infect larger quantities of cells for HisME53 purification.

2.3.4 Purification of His-ME53

Crude extract from Ac+HisME53-infected cells was mixed with 50% Ni-NTA slurry and loaded into a polypropylene column. The Coomassie blue staining of blots from eluates from the Ni-NTA column showed that imidazole concentrations from 150 mM began to elute a protein at the expected size of 53.5 kDa for HisME53 as seen by the major stained band in Figure 2.8. The optimal imidazole concentration for HisME53 elution was 250 mM as it eluted the greatest amount of protein based on visual density.

A higher concentration of imidazole (500 mM) released additional HisME53 bound to the column and there were fewer contaminating proteins eluted with HisME53 compared to the 250 mM eluate. The Western blot revealed His-tagged protein at the expected 53.5 kDa band as seen from the Coomassie blue stained gel (Figure 2.8A). An additional band at 100 kDa was visualized in eluates using 150 mM and 250 mM (Figure 2.8B) but was not easily detected by Coomassie blue staining. These 100 kDa bands could indicate dimer formation of ME53. The 100 kDa protein eluted from the column at lower concentrations of imidazole than monomer HisME53 did (Figure 2.8B). Lower binding affinity could be due to a dimerization domain close to the N-terminus and His tag, thereby decreasing the available surface area for the His tags to interact with the Ni2+ beads in the slurry. Dimer formation may be a by-product of overexpression. Preliminary yeast-2- hybrid work shows no evidence of ME53 dimer formation, and the putative dimers’

51

Figure 2.6: Map of his-tagged ME53 insertion into the commercial bacmid bMON14272 genome. Transposition occurs at the attTn7 site within a lacZ ORF at the polyhedrin locus. The donor plasmid, pFastBac HT B contains me53 cloned downstream of the polyhedrin promoter (green) and a TEV protease site (dark blue) fused to a 6x Histidine tag (light blue) for subsequent purification.

52

5E+07

4E+07

3E+07

/mL 50

TCID 2E+07

1E+07

0E+00 Ac+GFP Ac+HisME53

Figure 2.7: TCID50 titre of GFP-only and His-tagged ME53 viruses. The Ac+GFP virus titre is not significantly different from the Ac+HisME53 virus titre according to a Student’s two-tailed t-test p-value=0.47. The similar titres indicate that the addition of a 6x His N-terminal tag to ME53, and its overexpression does not affect BV production.

53

a b c d e f g h i A

B a b c d e f g h i

Figure 2.8: Validation of HisME53 purification. (A) Coomassie Brilliant Blue G250 stain of Ni-NTA column purification from Ac+HisME53 infected Sf9 cells. The lanes contain a) PageRuler Prestained Protein Ladder, with molecular sizes listed to the left, b) flow through from the column, c and d) wash buffer (20 mM imidazole) eluate, e) 100 mM imidazole elution buffer fraction, f) 150 mM imidazole elution buffer fraction, g) 250mM elution buffer fraction, h) 500 mM elution buffer fraction, Ia wash buffer (30 mM imidazole) eluate. (B) Western blot of transferred proteins from Ac+HisME53 protein extract purification. The lanes contain the same eluates as (A) and His-tagged proteins were detected using mouse anti-His as the primary antibody and HRP-conjugated rabbit anti-mouse as the secondary antibody.

54 concentration, possibly from overexpression, is much lower than the monomeric HisME53 as shown by its absence in the Coomassie blue stained gel. The detection of the 100 kDa protein via Western blot, and the absence of a corresponding heavy band on the

Coomassie stained gel could be the result of two His tags causing an amplification of fluorescence signal from the putative dimer.

Based on the gel analysis, ME53 was successfully overexpressed and purified using the baculovirus expression vector system. Successful infection and amplification demonstrate that overexpression of HisME53 does not cause cytotoxicity in Sf9 cells or impair BV production. Ac+HisME53 was amplified and BV infectivity was retained (Figure

+ 7 2.7). The average titre of Ac GFP (WT) after amplification was 2.66 x 10 TCID50/mL and

+ 7 was not significantly different from the Ac HisME53 titre of 1.04 x 10 TCID50/mL, that was averaged over two biological replicates. The absence of cytotoxicity in Ac+HisME53 infected cell indicates that HisME53 is not inhibitory to cellular or viral processes during very late times post infection.

Since a freeze thaw cycle of the eluate in 250 mM imidazole causes HisME53 to precipitate out of the solution, it is recommended to immediately dialyze the HisME53- containing eluate into TEV protease buffer. Further work should continue with the cleavage of the 6x His tag to ensure that it will not interfere with CD analysis to determine the influence of zinc on the structure of ME53.

Discussion

The lepidopteran baculovirus protein ME53 is highly conserved with several regions over 70% conserved in ME53 from all sequenced alpha and betabaculoviruses. Several of these domains are important for BV production as well as intracellular localization of

55

ME53. Four regions are of particular interest due to their validated or perceived functions: the NTS from aa 109-137, which facilitates ME53 nuclear translocation, as well as the putative ZnF-N residues 170-209, a non-characterized L region (residues 278-302) that is essential for BV production, and lastly, the putative ZnF-C region from residues 379-

399. The lack of non baculovirus homologues for ME53 makes function prediction challenging but some algorithms are capable of inferring disorder, secondary structure, binding regions, redox potential and solvent accessibility.

The NTS was used as a check-point for predictions as its role in baculovirus infection is well characterized. The NTS likely acts as a binding site for a viral chaperone protein that facilitates ME53 nuclear translocation during infection (Liu et al., 2016). The prediction algorithms suggest that the NTS is a binding domain, consistent with experimental evidence providing a threshold for the ANCHOR2 binding site prediction algorithm. Other putative binding sites include a region of the ZnF-N that correlates with a predicted -strand, and the C-terminus of the L region, which also correlates with a - strand. The ZnF-C region is not predicted to be a binding region, which is consistent with the experimental evidence that it non-significantly reduces BV output (de Jong et al.,

2009). The bioinformatic analyses predict that all of the conserved regions associated with BV production, the NTS, ZnF-N, and L, are also associated with binding regions; except for the ZnF-C domain. Although correlation does not equal causation, the results implicate ME53 as a multidomain protein with at least one binding partner, whose interaction with ME53 is essential for viral infection.

Zinc finger domains were initially classified as repetitive domains that coordinated zinc ions with two cysteine and two histidine residues to enable DNA binding and

56 recognition (Krishna et al., 2003). Now, the term zinc finger is used to describe any compact domain whose structure relies on the coordination of a zinc ion (Klug, 2010).

AcMNPV protein ME53 contains two putative C4 zinc finger domains predicted from the conserved cysteine residues whose spacing is consistent with zinc finger consensus sequences, C-X2-C-X33-C-X2-C for the ZnF-N, and C-X2-C-X14-C-X2-C for the ZnF-C. The objective of this section was to characterize the interaction of zinc with each of these regions; and use structural prediction software and data analysis tools to classify the domains preliminarily into fold groups.

The ZnF-N region is considered large for zinc finger domains as the whole motif stretches 39 aa with the two inner cysteine residues separated by 33 aa. The only structural fold group that includes large zinc finger domains is the treble clef finger fold group. The treble clef structure is characterized by a zinc knuckle followed by a loop, a - hairpin, and an -helix (Grishin, 2001). Structural prediction programs IUPred2A and

JPred4 indicate that the ZnF-N region is ordered with two short (6-8 aa) -strands and a short -helix beginning at the C-terminal cysteine residue similar to the positions of the secondary structures in the treble clef fold group (Grishin, 2001).

A peptide encompassing the putative ZnF-N domain was analyzed via CD spectroscopy in the absence and presence of zinc to determine if zinc contributes to the stabilization, or gain of structure, for the ZnF-N. A 5-fold molar excess of ZnCl2 to peptide resulted in a 5% increase in -strand structure for a total of 36% as well as a 3% increase in -helix structure to 7% and a 5% decrease in disorder to 60%. This is consistent with the JPred4 structure prediction of 32% -strand composition, 8% -helix and 59% irregular for the ZnF-N domain. The CD data indicates that zinc influences the secondary

57 structure of the ZnF-N peptide, supporting the prediction of a zinc finger domain in this region of ME53. The correlation between the JPred4 predicted secondary structure and the similar percentages of secondary structure predicted from the CD data, supports the classification of the N-terminal zinc finger in the treble clef zinc finger fold group.

The treble clef fold group of zinc fingers is further divided into 10 families of proteins with a wide variety of functions. Two immediate early baculovirus proteins, IE2 and PE38, fall within the first family of RING-finger-like proteins. Although the zinc-binding ability of

IE2 and PE38 to each bind two zinc ions has not been investigated, similar to ME53, they are assumed to bind zinc due to the RING finger consensus sequence C3HC4 (Prikhod’ko and Miller, 1998). The C4 zinc fingers, with >30 aa between cysteine pairs in the treble clef fold group, are the RPB10 protein subunit of RNA polymerase II and some members of the His-Me endonuclease family (Krishna, 2003). The majority of the His-Me endonuclease family members’ zinc binding residues have deteriorated but the general fold is still present. The exception is the T4 phage recombination endonuclease VII

(PDB:1en7) which is composed predominantly of -helices except for the presence of a

-hairpin that creates the treble clef fold within the larger DNA-binding domain. T4 endonuclease VII resolves Holliday junctions into duplex DNAs through single strand nicks to assist in the packaging of DNA into the phage head (Biertümpfel et al., 2007).

The RNA polymerase subunit protein RPB10 contains a treble clef zinc finger domain that is within a larger tertiary domain (Kaur and Subramanian, 2016). Mutation of any of the four cysteine residues results in a lethal phenotype in yeast as the zinc binding stabilizes the tertiary fold of the small protein (55-80 amino acids) (Mackereth et al., 2000). The large ZnF-N domain in ME53 may indicate that it is part of a larger tertiary structure and

58 the effect on BV production when this domain is deleted is due to the resulting instability of ME53. It is more likely a combination of structural integrity loss as well as the deletion of the putative binding domain identified by ANCHOR2 and indicates that the ZnF-N zinc finger structure is integral to the function of ME53.

The ZnF-C region of ME53 has the canonical consensus sequence of a zinc finger domain. The secondary structure was analyzed by CD using a peptide encompassing the

ZnF-C region (residues 377-401) to determine the effect of zinc on the peptide structure.

The addition of zinc resulted in a 6% increase in -strand structure to 37% which is lower than the JPred predicted 46% and much lower than the CAPITO Chou-Fasman prediction of 68% based on amino acid sequence. The NRMSD value for the secondary structure prediction of the ZnF-C peptide in the presence of zinc was high at 0.77, devaluing the secondary structure prediction based on the CD data. However, the distinct change in CD spectra of the peptide in the presence and absence of zinc indicates that the interaction of the ZnF-C domain with a zinc ion alters its secondary structure. The first identified zinc finger protein, TFIIIA, was analyzed using CD to confirm zinc binding by observing the structural change of a peptide encoding a single zinc finger from the protein (Frankel et al., 1987). The domain was initially referred to as a “zinc finger” due to the amino acid sequence indicating zinc binding potential but was confirmed as a zinc finger by using circular dichroism and cobalt spectrophotometry. The similar change in ZnF-C spectra observed with the addition of zinc strongly supports the use of zinc finger nomenclature when describing the ME53 ZnF-C domain. The canonical consensus sequence of the

ZnF-C and limited confidence of secondary structure prediction from the CD data, makes it difficult to predict the fold family for further classification of the domain. However, the

59

JPred prediction of three -strands may indicate a zinc ribbon fold. To absolutely confirm that the ZnF-C domain is a zinc finger domain, a high-resolution structure of completed

ME53 is required.

AcMNPV ME53 tagged with His was successfully expressed using BEVS and purified using Ni-NTA chromatography. The successful overexpression of ME53 indicated that ME53 is not inhibitory to the host or viral pathways essential for virulence or involved in BV morphogenesis. This was also seen with the AcMNPV per os infectivity factor P74, which when overexpressed did not impair BV production or infection in vivo (Faulkner et al., 1997). The overexpression of ME53 may impact BV production if it was overexpressed during earlier times post infection but the very late overexpression via the polyhedrin promoter indicates that ME53 is not inhibitory at late times post infection.

In summary, this study is the first to confirm that ME53 contains two zinc finger domains from aa 170-209 and 379-399, respectively which actually bind zinc. The overexpressed and purified ME53 provides a valuable reagent for further work to assess zinc binding of the whole protein, and to generate a full structure of ME53 to observe the zinc finger domains and classify them into fold families.

60

Chapter 3: The role of me53 domains in BV production and viral gene transcription

Introduction

Baculoviruses are large double stranded DNA viruses that infect hosts of the

Lepidopteran, Hymenopteran, and Dipteran orders. Baculoviruses produce two virion phenotypes during infection that are genetically identical but structurally different (Jehle et al., 2006). The first virion type, BV, is produced as early as 12 hpi until ~36 hpi, and is composed of a single nucleocapsid enveloped in a plasma membrane-derived lipid bilayer that mediates cell-to-cell transmission of the virus (Blissard and Rohrmann, 1990).

The second phenotype, ODV, is produced at very late times post infection ~24+ hpi and, in the case of AcMNPV, is composed of several nucleocapsids embedded in a polyhedrin crystalline matrix that is enveloped within the nucleus, and facilitates horizontal transmission of the virus (Rohrmann, 2013).

Immediate early protein ME53 is associated with the BVs and ODVs of AcMNPV and the BVs of Group II alphabaculovirus HearNPV (Hou et al., 2012). Deletion of me53 from the AcMNPV genome results in almost no BV production, implicating ME53 as a key component of BV production. Deletion analysis of several ME53 domains also indicates that the NTS and therefore the nuclear translocation of ME53 from 12 hpi is essential for optimal BV titres. Other domains involved in BV production include the L region (aa 278-

302) and the N-terminal region, aa 169-190, of the ZnF-N (aa 170-209). Deletion of the

ZnF-C domain reduces BV production at 72 hpi by 50%, so it may be important during early times post infection, or the ZnF-C may assist in protein-protein interactions at the plasma membrane for efficient BV assembly or egress. The influence of the entire ZnF-N domain or the role of the conserved cysteine residues, and their likely role in zinc

61 coordination on BV production has not been investigated and is included in this study. In addition, prior studies examined BV production after at least 24 hpt, allowing for a second viral life cycle to begin (Ono et al., 2015; Liu, 2015). This study focuses on the role of the

NTS, ZnF-N, L, and ZnF-C domains on BV production during the initial viral infection cycle up to 24 hpt to distinguish the temporal role of the domains during BV production.

Baculovirus gene expression is temporally regulated by conserved promoter consensus sequences that encourage gene expression via the host or viral RNA polymerase respectively (Jiang et al., 2006). The consensus sequences likely act as transcription factor binding sites for RNA polymerase recruitment. Viral transcription factors can enhance or reduce transcription by sequence-specific binding to promoter sequences and recruitment of the cellular RNA polymerase II, or viral RNA polymerase complexes for transcriptional activation (Li and Broyless, 1993). Transcriptional repression occurs when the transcription factor binds to a promoter sequence or enhancer sequence further upstream and prevents assembly of the RNA polymerase complex at the promoter (Lee et al., 1996). Other indirect methods to influence transcription include the activation or repression of cellular or viral transcription factors by transient binding interactions or phosphorylation (Sekimata et al., 2001; Balan et al.,

2016).

Early gene expression is regulated by specific cis-acting elements like the TATA and CAGT motifs assumed to be recognized by the host RNA polymerase (Jiang et al.,

2006). These immediate early genes then act as transactivators that regulate the expression of viral genes involved in downstream events such as DNA replication (Lu et al., 1997). Five immediate early genes including ie1 and me53 have validated or implied

62 transcriptional function. Deletion of ie1 results in widespread down-regulation of viral genes and is an essential transactivator for the stimulation of viral DNA replication; however, its transient expression results in apoptosis in Sf21 cells indicating that the temporal regulation of viral genes is crucial for virus infection (Prikhod’ko and Miller,

1999). ME53 is transcribed as early as 30 min post infection, is essential for BV production, and translocates to the nucleus between 12 and 18 hpi (Knebel-Morsdorf et al., 1996; de Jong et al., 2009; Liu et al., 2016).

ME53 also contains two C4 zinc finger domains that reduce BV production when deleted (Knebel-Morsdorf et al., 1993; de Jong, 2011). Zinc fingers were first recognized for their DNA binding capabilities and are well characterized as sequence-specific and general transcription factors, implicating the zinc finger domains of ME53 as trans-acting domains for the regulation of viral gene expression. The influence of ME53 on BV production, its immediate early expression and translocations to the nucleus during late times post infection, and the presence of two C4 zinc finger domains is the rationale for investigating the role of ME53 in viral gene transcription in the nucleus.

Methods

3.2.1 Bioinformatic analysis of ME53

To understand the evolutionary conservation of ME53 in alpha and betabculoviruses, a BLAST (https://blast.ncbi.nlm.nih.gov) search for AcMNPV ME53

(AA: AAA46718) in the Baculoviridae was conducted using the non-redundant protein sequence database and algorithm provided parameters. All sequences returned by

BLAST were used for the analysis of ME53 conservation through phylogeny and a representative alignment. The viral sequences returned are summarized in Table 3.1 with

63 the virus name, abbreviation, and accession number of the used sequence. Phylogenetic analysis, including alignment, was done using the MABL “One click” phylogeny analysis

(http://www.phylogeny.fr/simple_phylogeny.cgi) online platform. The alignment of representative sequences was completed using Clustal Omega

(https://www.ebi.ac.uk/Tools/msa/clustalo/).

3.2.2 Me53 knockout bacmid construction

The commercial bacmid bMON14272 (Invitrogen) containing the whole AcMNPV genome, except for the polyhedrin gene, was maintained in E. coli DH10 cells. The me53 ORF was deleted and replaced with the chloramphenicol acetyl transferase (cat) gene under control of a bacterial promoter by Dr. Emine Ozsahin (University of Guelph).

The me53 bacmid was transposed with gfp under the control of the me53 promoter and the SV40 polyadenylation sequence from the pFACTproMGFP donor plasmid. The plasmid included a gentamicin resistance gene under the control of a bacterial promoter, the AcMNPV polyhedrin gene downstream of its native promoter, and a Tn7 cassette, which was used to transpose gfp and polyhedrin into the attTn7 transposition site to generate AcME53GFP. The bacmid was sequenced to ensure that the me53

ORF was deleted as designed. Generation of bacmids containing GFP-tagged ME53 internal deletions and site-directed mutagenesis

Recombinant bacmids were constructed using the me53 bacmid and the pFACT plasmid (de Jong, 2011). The gfp open reading frame (ORF) was amplified from the pGEM+T-GFP plasmid using primers GFP-F

5’CTGCAGGTGAGCAAGGGCGAGGAGCTG-3’ (PstI site is underlined), and GFP-R 5’-

GATATCTTACTTGTACAGCTCGTCCAT-3’ (EcoRV site is underlined). The fragment

64

Table 3.1: Baculoviruses and their abbreviations and accession numbers used in the ME53 phylogenetic analysis Virus name Abbreviation Accession number Autographa californica nucleopolyhedrovirus AcMNPV AAA46718 Plutella xylostella multiple nucleopolyhedrovirus PlxyNPV ABE68522 Rachiplusia ou multiple nucleopolyhedrovirus RoMNPV AAN28036 Bombyx mandarina nucleopolyhedrovirus BomaNPV AFO10091 Bombyx mori nucleopolyhedrovirus BmNPV AFN21095 Thysanoplusia orichalcea nucleopolyhedrovirus ThorNPV YP_009255275 Maruca vitrata nucleopolyhedrovirus MaviNPV YP_950837 Catopsilia Pomona nucleopolyhedrovirus CapoNPV YP_009255275 Lonomia obliqua multiple nucleopolyhedrovirus LoobMNPV AKN81021 Oxyplax ochracea nucleopolyhedrovirus OxocNPV AVA31115 Antheraea pernyi nucleopolyhedrovirus AnpeNPV BAX08787 Samia ricini nucleopolyhedrovirus SariNPV BBD51082 Philosamia Cynthia ricini nucleopolyhedrovirus PhcyNPV AFY62823 Cyclophragma undans nucleopolyhedrovirus CyunNPV AOT85487 Choristoneura muriana nucleopolyhedrovirus ChmuNPV AHD25504 Hyphantria cunea nucleopolyhedrovirus HycuNPV BAE72306 Spilosoma obliqua nucleopolyhedrovirus SpobNPV AUR45047 Orgyia pseudotsugata nucleopolyhedrovirus OpMNPV AAC59136 Condylorrhiza vestigialis multiple nucleopolyhedrovirus CoveMNPV AJD09182 Choristoneura fumiferana multiple nucleopolyhedrovirus CfMNPV AAP29907 Dasychira pudibunda nucleopolyhedrovirus DapuNPV AKR14095 Choristoneura occidentalis alphabaculovirus ChocNPV AGR56906 Choristoneura rosaceana nucleopolyhedrovirus ChroNPV AGR57054 Anticarsia gemmatalis multiple nucleopolyhedrovirus AgMNPV ALR72342 Epiphyas postvittana nucleopolyhedrovirus EppoNPV AAK85686 Agrotis segetum nucleopolyhedrovirus A AgseNPV-A AAZ38173

65

Agrotis ipsilon multiple nucleopolyhedrovirus AgipMNPV ACI28712 Agrotis segetum nucleopolyhedrovirus B AgseNPV-B AIZ48566 Spodoptera exigua multiple nucleopolyhedrovirus SeMNPV CDG72350 Peridroma sp. Nucleopolyhedrovirus PespNPV AIE47739 Spodoptera litura nucleopolyhedrovirus SltNPV ACI47377 Chrysodeixis chalcites nucleopolyhedrovirus ChchNPV AAY83939 Trichoplusia ni single nucleopolyhedrovirus TnSNPV AAS82879 Lymantria dispar multiple nucleopolyhedrovirus LdMNPV AJR20294 Lymantria xylina nucleopolyhedrovirus LyxyMNPV ADD73730 Spodoptera frugiperda multiple nucleopolyhedrovirus SfMNPV ACA02567 Malacosoma sp. alphabaculovirus MaspNPV ANW12314 Mythimna unipuncta nucleopolyhedrovirus MyunNPV AUV65269 Chrysodeixis includens nucleopolyhedrovirus ChinNPV AOL56446 Apocheima cinerarium nucleopolyhedrovirus ApciNPV ADB84468 Pseudoplusia includens nucleopolyhedrovirus PiNPV AJD80700 Malacosoma neustria nucleopolyhedrovirus ManeNPV AUF91545 Cryptophlebia peltastica nucleopolyhedrovirus CrpeNPV AXS67682 Leucania separata nucleopolyhedrovirus LsNPV AAR28791 Ectropis obliqua nucleopolyhedrovirus EcobNPV ABI35697 Hemileuca sp. nucleopolyhedrovirus HespNPV AGR56760 Mamestra configurata nucleopolyhedrovirus A MacoNPV-A AAQ11027 Mamestra configurata nucleopolyhedrovirus B MacoNPV-B AAM94994 Clanis bilineata nucleopolyhedrovirus ClbiNPV ABF47357 Mamestra brassicae multiple nucleopolyhedrovirus MabrNPV AFL64858 Helicoverpa armigera multiple nucleopolyhedrovirus HearMNPV ACH88529 Orgyia leucostigma nucleopolyhedrovirus OrleNPV ABY65740 Helicoverpa zea single nucleopolyhedrovirus HzSNPV AAL56161 Helicoverpa assulta nucleopolyhedrovirus HasNPV AXR98007

66

Euproctis pseudoconspersa nucleopolyhedrovirus EupsNPV ACO53465 Adoxophyes orana nucleopolyhedrovirus AdorNPV ACF05317 Lambdina fiscellaria nucleopolyhedrovirus LafiNPV AKC91748 Buzura suppressaria nucleopolyhedrovirus BusuNPV AHH82611 Hyposidra talaca nucleopolyhedrovirus HytaNPV AWW14385 Perigonia lusca single nucleopolyhedrovirus PeluSNPV AKN80620 Operophtera brumata nucleopolyhedrovirus OpbuNPV AUA60251 Spodoptera littoralis nucleopolyhedrovirus SpliNPV AGE89878 Adoxophyes honmai nucleopolyhedrovirus AdhoNPV BAC67277 Sucra jujuba nucleopolyhedrovirus SujuNPV AIU41255 Spodoptera litura granulovirus SpliGV ABQ52077 Plutella xylostella granulovirus PxGV ANY57639 Urbanus proteus nucleopolyhedrovirus UrprNPV AKR17337 Cryptophlebia leucotreta granulovirus CrleGV AUF82002 Cnaphalocrocis medinalis granulovirus CnmeGV ALN42066 Diatraea saccharalis granulovirus DisaGV AKN80765 Spodoptera frugiperda granulovirus SfGV AJK91806 Epinotia aporema granulovirus EpapGV AER41558 Cydia pomonella granulovirus CpGV AAK70803 Mocis latipes granulovirus MolaGV AKR17478 Artogeia rapae granulovirus ArGV ACZ63606 Phthorimaea operculella granulovirus PhopGV AAM70328 Pieris rapae granulovirus PrGV AGS18874 Mythimna unipuncta granulovirus MyunGV YP_009345870 Xestia c-nigrum granulovirus XcGV AAF05294 Pseudalatia unipuncta granulovirus PuGV ACH69532 Trichoplusia ni granulovirus TnGV AOW41509 Helicoverpa armigera granulovirus HearGV ABY47869

67 was cloned into pBluescriptSV40, which contains an SV40 polyadenylation sequence, using PstI and EcoRV generating pBluescriptGFPSV40. The me53 promoter and ORF was amplified from the AcMNPV bacmid using primers me53proFw 5’-

CGAGCTCAGCGTGTGCGCCGGAG-3’ (SacI site is underlined), and me53Rv 5’-

AACTGCAGGACATTGTTATTTACAATATTAAT-3’ (PstI site is underlined) and cloned into pBluescriptGFPSV40 using SacI and PstI generating pBlueproME53:GFPSV40. The entire me53 gene including gfp and the SV40 sequence was subcloned into the pFACT plasmid using SacI and XhoI generating pFACTproME53:GFPSV40. Internal deletions of me53 in the pFACTproME53:GFPSV40 plasmid were carried out using inverse PCR or a two-step mutagenesis PCR with the phosphorylated primer sets listed in Table 3.2 to generate the corresponding plasmids. Briefly, the PCR product was purified according to the manufacturers’ instructions (Wizard SV Gel and PCR Cleanup Kit) and treated with 2

L of DpnI (Thermo Fisher) in 50 L for 15 minutes at 37°C. The Dpn1 digested product was cleaned a second time (Wizard SV Gel and PCR Cleanup Kit) and 20 ng of DNA was ligated overnight at 4°C, using T4 Ligase (Thermo Fisher). The Tn7 cassette from pFACTproME53:GFPSV40, and other modified plasmids, was transferred to the attTn7 transposition site in the me53 bacmid, as described in the Invitrogen Bac-to-Bac expression manual, to generate AcME53-ME53:GFP (WT) and other recombinant bacmids. All plasmids and recombinant bacmids were verified by sequencing and GFP expression respectively.

3.2.3 Transfections for growth curve analysis

Sf9 cells were seeded at 2 x 105 cells/well in 12-well plates and allowed to adhere for at least one hour. One g of purified bacmid DNA in 50 L Grace’s SFM (Thermo

68

Fisher) was combined with 2 L Cellfectin II (Invitrogen) in 50 L Grace’s SFM, gently mixed, and incubated for 30 minutes. The Hink’s + FBS medium was removed from the

Sf9 cells and replaced with 500 L Grace’s SFM for 30 minutes to rinse the cells. The

Cellfectin II and bacmid DNA solution was increased to 400 L using Grace’s SFM and added to a well of Sf9 cells after the Grace’s SFM was removed. Cells and bacmid DNA were incubated for 5 h and then the medium, bacmid, and Cellfectin II solution were removed from the cells and replaced with 1 mLof Hink’s + FBS supplemented with 1X anti-mycotic anti-biotic (Gibco, 15240-062). Cell culture supernatant was collected at 0,

6, 12, 18, or 24 hours post transfection (hpt) and placed in 1.5 mLcentrifuge tubes for centrifugation at 1000 x g for 5 min to pellet any cells. The centrifuged supernatant was removed and used for growth curve analysis.

3.2.4 Determination of viral titre

Viral titre was determined using the TCID50 end-point dilution method (Reed and

Muench, 1983). Briefly, virus supernatant was serially diluted from 100-10-9 and 100 L of each dilution was added to eight wells of a 96-well plate seeded with 100 L per well of Sf9 cells at a density of 1.0 x 105 cells/ml. Eight wells of the infected plates were left uninfected as a negative control. The 96-well plates were incubated at 27°C for 7 days before they were scored for the presence of virus based on obvious infection signaled by occlusion bodies. Titre was calculated by the Reed-Muench TCID50/mL method.

69

Table 3.2: Primers for internal deletions and mutations of me53 in pFACTproME53:GFP. All primers are 5’- phosphorylated. Plasmi Gene Primer Name Sequence 5’-3’ d Name P-NTS Del Fw GGCTACATAAACAGCGAAGAT NTS ME53 109-137  P-NTS Del Rv GCATCGCGAATCGCTCAT

P-ZnF-N Del Fw TATCTATATCACAACGTTCCA ZnF-N ME53 170-209  P-ZnF-N Del Rv GCATCGCGAATCGCTCAT

P-L Del Fw ACCAATATAGTGTTGCAGCG L ME53 278-302  P-L Del Rv GTGATTTCGAATTATGCGCCG P-ZnF-C Del Fw GGGTTTACCAACGTTTATCATTTTCC ZnF-C ME53 379-399  P-ZnF-C Del Rv GCAATAGTTGTTAGTTGTAGGTTTCAAA P-ZnF-N Del Fw TATCTATATCACAACGTTCCA ME53170-209, P-ZnF-N Del Rv GCATCGCGAATCGCTCAT ZnF 379-399 P-ZnF-C Del Fw GGGTTTACCAACGTTTATCATTTTCC P-ZnF-C Del Rv GCAATAGTTGTTAGTTGTAGGTTTCAAA P-ZnF-N C170A C173A GATTCGCGAGCCACCACTGCCAATTATAGATTCAAAGACAA Fw C P-ZnF-N C170A C173A GCTCATGTGCTCCATAGTCTTC ME53C170A, ZnF-N Rv CATAGACATCTGCGCTCAAAAAGCCTATCTATATCACAACGT C173A, C206A, 4C→4A P-ZnF-N C206A C209A TCC C209A Fw CGGTCAGGATCGTCCAACGGCTTTTCGA P-ZnF-N C206A C209A Rv

70

P-ZnF-C C379A C382A CTAACAACTATGCCAAATTGGCTAAAAAAACTAAACTGTATT Fw ATAAAAATCCCG GGGATTTTTATAATACAGTTTAGTTTTTTTACAC P-ZnF-C C379A C382A GTGTTATATGCTACCAAAGCCGGGTTTACCAACGTTTATCAT ME53C379A, ZnF-C Rv TTTCCT C382A, C396A, 4C→4A P-ZnF-C C396A C399A GGGATTTTTATAATACAGTTTAGTTTTTTTACAC C399A Fw

P-ZnF-C C396A C399A Rv

71

3.2.5 Viral gene selection for qRT-PCR analysis

Viral genes were selected based on the promoter consensus sequences found within the 300 base pairs upstream of the translation start site. The consensus sequences included were the tetranucleotide repeats: TATA, CAGT, CATT, CAAT, TCGT, GTAG,

AGTC, TAGG, TAAG, and the heptanucleotide sequence TATAAGG (Xing et al., 2005).

The TATA, CAGT, CATT, and CAAT sequences upstream of each viral gene are considered early motifs as they are recognized by the host RNA polymerase II complex.

Tetranucleotide sets TCGT, GTAG, and AGTC are found upstream of both early and late genes and are surmised to be expressed by the host RNA Polymerase II and viral RNA polymerase complexes respectively. The TAGG, TAAG, and TATAAGG sequences are considered strong late promoter consensus sequences frequently found upstream of highly expressed late genes including polyhedrin (Xing et al., 2005). The promoter sequences were also analyzed using Clustal Omega to generate pairwise alignments and a percent identity matrix that compared the number of identical residues to sequence length. The pairwise alignments indicate the overall similarity of the promoter sequences and were used to ensure the promoter sequences did not have >50% similarity so that the consensus sequences, and not overall similarity, are the cis-acting elements potentially influencing viral gene expression.

Viral genes were also selected for their temporal expression patterns and include an immediate early gene, ie-1, expressed immediately following viral infection, and odv- e25, a gene encoding the occlusion derived protein involved in ODV envelopment and produced at very late times (24+ hpi). The other genes selected using the criteria

72 described dictatum erat include the viral RNA polymerase subunit genes lef-9 and lef-8, late gene vp80, and gp64, which encodes the BV major envelope glycoprotein.

3.2.6 Transfection for transcriptional analysis

Bacmid DNA was isolated from transformed E. coli DH10 cells cultured overnight in 3 mL LB medium. The DNA concentration was determined using a NanoDrop ND-1000 and those samples giving A260/A280 ratios between 1.8 and 2.0 were accepted for further analysis. Two mL of Sf9 cells at a density of 7.5 x 105 cells/mLwere seeded in 60 mm dishes and allowed to adhere for 1 h before they were transfected with 5 g of bacmid

DNA using 8 L of Cellfectin II according to the manufacturer’s protocol as previously described. The cells were incubated with the DNA/Cellfectin II solution for 5 hours before the solution was removed and replaced with 2 mLof Hink’s +FBS supplemented with 1X anti-mycotic and anti-biotic (Gibco).

3.2.7 RNA isolation from transfected cells

Total RNA from transfected Sf9 cells was isolated at 12 and 18 hours post transfection using the Qiagen RNeasy Mini Kit following the manufacturer’s instructions.

Cell lysate was homogenized by vortexing and eluted with the RNase-free water provided.

RNA concentration was measured using a NanoDrop ND-1000 and 6 g total RNA was digested with RNase free DNase (Thermo Fisher) in a 30 L reaction for 1 hour at 37°C to remove the remaining DNA. After 1 h, 2 L of 50 mM EDTA was added and the sample was incubated at 65°C for 20 min to inactivate the DNase. The RNA was stored at -20°C until needed for downstream work.

73

3.2.8 Generation of cDNA using reverse transcription

The RNA concentration, after DNase digestion, was measured prior to reverse transcription. Quanta Biosciences’ qScript cDNA Supermix was used to synthesize cDNA from 1 g DNase-treated RNA in a 20 L reaction. The reverse transcription reaction was incubated in a MyCycler Thermocycler (BioRad) for 5 minutes at 25°C, 30 minutes at

42°C, 5 minutes at 85°C and held at 4°C. The concentration of cDNA was measured after reverse transcription and cDNA was diluted to 50 ng/L to ensure equal amounts of cDNA were used for qPCR.

3.2.9 Primer amplification efficiency assay

The amplification efficiency of primers used for viral transcript analysis must be within 90-110% of the primers used for the internal control (Taylor et al., 2010).

Additionally, the reactions must amplify product within 10 to 30 cycles for the highest sensitivity, therefore input concentration of cDNA template must be optimized. To accomplish this, qPCR reactions with 10 L SYBR Green I (Bioline), 0.5 M forward and reverse primers, and 250 ng, 100 ng, or 50 ng of cDNA template were set up. The PCR reaction cycle was as follows: 95°C for 2 min once, 40 cycles of 95°C for 30 seconds,

60°C for 10s, 72°C for 20s followed by 83°C for 3 sec for data acquisition using the

StepOnePlus Real-Time PCR System from ABI. A standard curve for each primer set was established by plotting the log10 of the cDNA concentration against the CT value of the primer sets. The correlation coefficient (r2) threshold of the standard curve was >0.980 indicating an amplification efficiency within 90-110% for a primer set (Taylor et al., 2010).

74

Table 3.3: Primer pair sequences used in the qRT-PCR assay for viral gene expression analysis. Gene Primer Name Sequence (5’-3’) ie1 ie1-F ATCGCCCAGTTCTGCTTATC ie1-R TTCGTCCAGCTTCCGTTTAG lef-8 lef-8-F CTACTCTGTGGCGGTAAACAA lef-8-R GTGCGGAATGTACGGATCTT lef-9 lef-9-F ACGCGTCATCCCAACATTAG lef-9-R GTGTTAGCGCCGGGAAATA vp80 vp80-F GCAAACGGTGGATGTGATTG vp80-R CTCCGCTTCCTGTTCTCTTT gp64 gp64-F GTGAAGCGGCAGAATAACAATC gp64-R CCTCTGTGTACTTGGCTCTAAC odv-e25 odv-e25-F AACAAGGTGGGCACTAACA odv-e25-R GGGTCGATCAAACACACAATC

75

3.2.10 Viral transcript analysis using qRT-PCR

Viral genes chosen for RT-qPCR analysis were monitored at 12 and 18 hpt in cells transfected with the bacmids generated in section 3.2.3. Transcript levels were analyzed from two biological replicates with three technical replicates each and compared to

Ac+proGFP transcript levels. The cellular 28S rRNA was used as an internal reference gene to control for RNA extraction efficiency (Salem et al., 2014). The qPCR reactions were carried out with 50 ng of cDNA, 10 L SYBR Green No Rox (Bioline), and 0.5 M of forward and reverse primers in 20 L the primer pairs are listed in Table 3.3. The reactions were completed in 96-well plates in a StepOnePlus Real-Time PCR System from ABI under the same parameters as 3.2.10. Data were analyzed using the Ct analysis method to compare the difference in Ct value between viral transcripts and 28S rRNA of Ac+proGFP bacmid transfections and ME53 variant bacmid transfections. The fold-change in viral transcript levels was determined by comparing viral gene transcript levels that have been normalized to the 28s rRNA gene within one sample to the WT sample using the equation:

Fold-change = 2-Ct

CT=(Cptarget-Cpreference)experimental – (Cptarget-Cpreference)control

Results

3.3.1 Identification of conserved regions targeted for deletion/mutagenesis

Bioinformatic analyses are valuable tools in genetics to classify genes according to sequence homology, identify key residues from their evolutionary conservation, and advance research by generating sequence-based hypotheses. Viruses are model systems to study genetic conservation at the whole genome level due to their small

76 genomes, short generation time, and high mutation rates, with gene and residue conservation indicative of functional importance (Duffy et al., 2008). The AcMNPV gene, me53, is essential for BV production and is conserved in all alpha- and betabaculoviruses sequenced to date (Figure 3.1). The conservation of ME53 during baculovirus evolution separated into three distinct clades. The first clade was composed of Group I alphabaculoviruses, including AcMNPV, which contain eleven unique open reading frames not found in the second clade comprised of Group II alphabaculoviruses. The shorter branch lengths of Group I alphabaculovirus genomes indicated closer relationships between viruses of that clade compared to the longer branch lengths seen in the Group II alphabaculovirus clade which consequently indicated more divergence between sequences. This is likely due to the more recent separation of alphabaculoviruses into two groups, with Group I considered the newest clade in baculovirus evolution history (Jiang et al., 2006). The distinct separation of ME53 between the two clades indicates that ME53 predates the evolutionary divergence that resulted in

11 additional ORFs in Group I alphabaculoviruses (Miele et al., 2011). The third clade comprised the betabaculoviruses, which contain an N-terminal truncated version of

ME53. Betabaculoviruses lack the first 110 amino acids of the ME53 N-terminus compared to alphabaculovirus ME53. ME53 of five viruses, OpbuNPV, SpliNPV, LsNPV,

HzSNPV, and HasNPV did not group with any of the three clades although they are all classified as Group II alphabaculoviruses (Huang and Levin, 2001; Broadly et al., 2016;

Xiao and Qi, 2007; van Beek and Davis, 2007). The use of ME53’s amino acid sequence for phylogenetic analysis versus complete genomes, or several ORFs, was the likely

77 reason for their separation. The conserved regions of ME53 considered in this thesis are conserved in the five non-grouped Group II alphabaculoviruses.

Several regions of me53 share over 50% conservation which is indicative of functional importance. A representative alignment using five ME53 aa sequences from

Group I alphabaculovirus AcMNPV, Group II alphabaculoviruses TnSNPV and LdMNPV, and betabaculoviruses PxGV and CpGV is shown in Figure 3.2. The alignment highlighted several regions of ME53 determined, or proposed, to affect BV production.

The previously identified NTS from resideus 109-137 of AcMNPV ME53, in the blue box, showed that 11% of NTS residues were fully conserved while an additional 29% of residues were highly conserved or similar. The ZnF-N from residues 170-209 (AcMNPV

ME53), showed 100% conservation of the cysteine residue locations at aa 170, 173, 206 and 209, and phenylalanine at aa 177 as well as a conserved region of hydrophobic residues from residues 187-190. The conserved L region between the two zinc finger domains from residues 278-302, shown in the purple box, is important for BV production at ≥ 72 hpi (de Jong, 2011). Of residues in this region 12.5% (K279, I285, and L290) were

100% conserved, and 21% were highly conserved suggesting it may be a conserved functional region. The IUPred ANCHOR2 output from section 2.3.1 predicted a binding domain within this region, supporting a putative functional role. The second C4 zinc finger domain, ZnF-C, from residues 379-399, showed 100% conservation of the cysteine residue locations (at aa 379, 382, 396 and 399), and a six-aa non-polar region (aa 393-

398) close to the C-terminal cysteines that is highly conserved across all genera.

78

Figure 3.1: Maximum likelihood phylogenetic tree of ME53 sequences. Maximum likelihood phylogenetic tree of ME53 sequences in all alpha and betabaculovirus genomes sequenced to date. The tree was assembled using PhyML with the WAG model of amino acids substitution based on the sequence alignment by MUSCLE. The betabaculoviruses are shown in green, group II alphabaculoviruses in red, and Group I alphabaculoviruses in blue, including AcMNPV ME53 which is circled. The viruses shown in black do not group well with any of the three families based only on the amino acid sequence of ME53.

79

NTS

ZnF-N

L

ZnF-C

Figure 3.2: Alignment of ME53 sequences. Alignment of ME53 sequences from a group I alphabaculovirus (AcMNPV), two group II alphabaculoviruses (TnSNPV and LdMNPV), and two granuloviruses (PxGV and CpGV). The boxed areas correspond to the NTS (blue), the putative ZnF-N (red), a conserved region important for BV production (purple), and the putative ZnF-C (green). Clustal Omega was used for the sequence alignment with default parameters and the graphic was generated by ESPript.

80

Residue conservation indicating functional significance was the rationale for selecting regions of ME53 to delete or mutate for domain characterization in BV production. The regions chosen for deletion or mutation from me53 included the NTS, the

ZnF-N, L region, the ZnF-C, and both C4 zinc fingers. Also included for their level of conservation and implied coordination of zinc ions, are the respective cysteine residues from each zinc finger domain. All four of the ZnF-N cysteine residues were mutated to alanine, C170A, C173A, C206A, and C209A, as were all four of the ZnF-C cysteines,

C379A, C382A, C396A, and C399A, respectively.

3.3.2 Generation of bacmids with ME53 deletion/mutation variants

Bacmids were manipulated in order to study the effect of ME53 deletions and mutations on AcMNPV infectivity and BV production. Essentially me53 was cloned into a donor plasmid (pFACT) containing a C-terminal GFP tag to confirm ME53 expression and follow replication. In-frame deletions of the NTS, ZnF-N, L, ZnF-C, and both zinc fingers, were done via inverse PCR of the whole plasmid, along with site-directed mutagenesis for the cysteine residues of each zinc finger domain, and confirmed by sequencing. The GFP tag was used to confirm transposition of me53 into an me53-null bacmid (Acme53) by monitoring GFP expression in transfected cells. A total of nine

Acme53 bacmids were generated with the appropriate ME53 deletions and site- directed mutagenesis of the cysteine residues of each zinc finger (Figure 3.3).

3.3.3 Virus replication of AcMNPV with conserved domain deletions in me53

To determine the effect of ME53 domain deletion and mutagenesis on BV production in the initial viral infection, the nine bacmids ut supra were transfected separately into monolayers of Sf9 cells and the supernatant was collected every 6 hs from

81

WT

me53

NTS

ZnF-N

L

ZnF-C

ZnF

ZnF-N 4C→4A

ZnF-C 4C→4A

Figure 3.3: Schematics of bacmids generated for growth curve and RT-qPCR analysis. Schematics of GFP-tagged ME53 internal deletions and mutations transposed into an Acme53 bacmid. The me53 promoter (proM) gfp (purple arrow) is fused to the C- terminus of me53.

82

0 to 24 hpt inclusive. Previous work determined that me53 is necessary for BV production by determining virus titre after 24 hpt. ME53 translocates to the nucleus during infection, and later localizes to foci at the plasma membrane suggesting at least two discrete functions, one in the nucleus another at the plasma membrane. Functional analysis of

ME53 domains in AcMNPV infection requires transfection for infection initiation, as deletion of me53 results in limited BV production and virus cannot be amplified. Care was taken to ensure the same amount of bacmid DNA was used for transfection for each of the different constructs by accurately measuring DNA amounts and performing 4 biological replicates to account for any variation. The infectious virus titre of supernatant collected from transfected cells was calculated as TCID50 (Reed and Muench, 1983). BV titre for each bacmid over a 24-hour period was measured (Figure 3.4).

Wildtype AcMNPV bacmid (Ac+ME53:GFP) transfection increased BV production

4 from 6 to 24 hpt with a maximum titre of 3.18 x 10 TCID50/ml. A 6-fold increase in BV production was seen from 18 to 24 hpt, correlating to the earliest time of translocation of

ME53 to the nucleus. Acme53 transfected cells show very limited BV production through the initial stage of infection. At 24 hpt, Acme53 transfected cells produced a titre of 3.44

4 TCID50/mLcompared to the 3.18 x 10 TCID50/mLtitre produced by WT transfected cells.

A significant difference in BV production between WT and Acme53 transfected cells was

3 4 seen at 12, 18, and 24 hpt, with WT titers of 18.4, 4.18 x 10 , and 3.18 x 10 TCID50/ml, and Acme53 titres of 2.07, 2.52, and 3.44 TCID50/ml, respectively. Acme53NTS transfected cells showed a production pattern similar to Acme53 from 0 to 12 hpt.

Acme53NTS titre increased slightly from 18 to 24 hpt compared to Acme53 and produced a maximum BV titre of 10.6 TCID50/mLat 24 hpt. Transfection of cells with the

83

Acme53ZnF-N bacmid showed a similar inhibition of BV production from 0 to 12 hpt as transfection with Acme53 and Acme53NTS. However, by 18 hpt Acme53ZnF-N BV titre was 52.3 TCID50/ml, or 20.7-fold higher than Acme53, but 80.1-fold lower than WT

2 transfected cells, and by 24 hpt the titer of ZnF-N was 1.31 x 10 TCID50/ml, or about

37.8-fold higher than Acme53 and 243-fold lower than WT titre. The difference in BV production between Acme53 and Acme53ZnF-N was not significant at any time points post transfection but was significant compared to WT transfected cells at 12, 18, and 24 hpt (p<0.05). Deletion of the conserved L region (residues 278-302) resulted in BV production similar to Acme53NTS and Acme53 with a maximum titre of 16.6 TCID50/mL at 24 hpt. The BV titre from transfected cells was significantly different from the BV titre of WT transfected cells at 12, 18, and 24 hpt. Transfection of Sf9 cells with Acme53ZnF-

C resulted in a delay of BV production between 0 and 12 hpt. Acme53ZnF-C produced

BV titres similar to Acme53 transfected cells from 0 – 12 hpt but BV production increased

2 4 from 2.61 TCID50/mL at 12 hpt to 1.42 x 10 TCID50/mL at 18 hpt then to 1.70 x 10

TCID50/mLat 24 hpt for a 54.5-fold increase in titre from 12 – 18 hpt and a 119-fold increase from 18 – 24 hpt, respectively. The viral growth kinetics of Acme53ZnF-C between 12 and 18 hpt were significantly lower than WT, but at 24 hpt the Acme53ZnF-

C titre was significantly higher than all other constructs excluding WT (p<0.05). The BV titre of Acme53ZnF (with both zinc fingers deleted) transfected cells mimicked that of

Acme53 at all time points and was lower than both Acme53ZnF-N and Acme53ZnF-

C transfected cells. At 24 hpt the titer of Acme53ZnF was 17.6 TCID50/mL. The difference in BV titre between Acme53ZnF and Acme53ZnF-N was not significant at

84

1E+05

Ac+ △ME53 △NTS 1E+04 △ZnF-N △ZnF-C △L △△ZnF

/mL) 1E+03

50 * * BV Titre (TCID Titre BV 1E+02

1E+01 *

1E+00 0 6 12 18 24 Hours post transfection

Figure 3.4 The effect of ME53 domain deletions on viral growth kinetics. The effect of conserved domain deletions on BV growth from 0-24 hpt at 6 hour intervals. The BV titre is shown in TCID50 units/mLon a logarithmic scale with the standard deviation bars representing four biological replicates. The WT growth curve is shown in black squares followed by me53 deletion in open triangles, me53NTS in open circles, me53ZnF-N in open squares, me53ZnF-C in x-markers, me53L in asterisks markers, and me53ZnF in black triangles. Significance is indicated by a black asterisk. The data points above the asterisk are significantly different (p<0.05, student’s two-tailed paired t-test) than those below.

85 any time points compared to the significant 960-fold difference at 24 hpt between the BV

4 titres of Acme53ZnF at 17.6 TCID50/mL and Acme53ZnF-C at 1.70 x 10 TCID50/ml, respectively. These data indicate that all of the investigated regions of ME53, with the exception of the ZnF-C, are required for optimal BV production from 12 -24 hpt, when

ME53 localizes to the nucleus. The ZnF-C is required for early BV production from 0 – 18 hpt a time when ME53 is localized in the cytoplasm and begins to translocate to the nucleus, suggesting that the ZnF-C has a cytoplasmic role which can be overcome later in infection.

3.3.4 The role of zinc coordination by ME53 in BV production

While it is clear that individual or simultaneous deletion of the zinc finger domains in ME53 adversely affects virus replication it is not clear if this is because they function as zinc fingers or simply represents an otherwise important region of ME53. Zinc finger domains rely on the coordination of zinc for their secondary structure and exposure of residues involved in the domains’ function, and deletion or mutation of these residues can result in limited or non-functioning proteins. The ZnF-N and ZnF-C domains of ME53 are highly conserved and are capable of interacting with zinc ions likely via the absolutely conserved cysteine residues that flank the domains (as shown in Chapter 2.3.2). Deletion of each domain resulted in decreased BV production with deletion of the ZnF-N region causing a similar decrease in BV production as seen with deletion of the entire me53. In contrast, removal of the ZnF-C resulted in only a delay in early BV production, but from

12 to 24 hpt when ME53 translocates to the nucleus and BV production increases,

Acme53ZnF-C BV production increased to ~50% of WT. When both zinc coordinating regions were simultaneously deleted, a negative synergistic effect (i.e. the decrease in

86 titer of the double knock out was greater than the knockout of either ZnF alone) was seen on BV production.

To better characterize the roles of the putative zinc-binding cysteine residues in

BV production, the cysteine residues of each zinc coordinating region were mutated to alanine residues and transfected into Sf9 cells and assayed ut supra. Ironically, the deletion of the ZnF-N had less of an effect on virus growth compared to alanine mutagenesis of the ZnF-N cysteine residues (Figure 3.5A). Putative impairment of zinc binding via alanine mutagenesis of the cysteine residues in ZnF-N resulted in a BV production curve similar to that for Acme53. From 0 – 12 hpt, there was no significant difference between the Acme53ZnF-N and Acme53ZnF-N4C→4A transfected cells. At

18 hpt, the 52.2 TCID50/mLtitre of Acme53ZnF-N was 24.4-fold higher than the 2.14

TCID50/mL titre of Acme53ZnF-N4C→4A resulting in a p-value of 0.08 due to the larger standard deviation of the Acme53ZnF-N titre. A similar difference was seen at 24 hpt

2 with the Acme53ZnF-N BV titre of 1.3 x 10 TCID50/mLbeing 12.5-fold higher than the

Acme53ZnF-N4C→4A BV titre of 10.5 TCID50/ml, although the difference was not significant (p=0.17). Mutagenesis of the cysteine residues did not result in a significant decrease in BV production compared to full domain deletion but there was a 10 to 20-fold decrease at late times post transfection, when ME53 localizes to the nucleus, and begins to form foci at the plasma membrane suggesting a comparable function at this time.

Likewise, deletion of the ZnF-C had less of an effect on virus growth compared to alanine mutagenesis of the ZnF-C cysteine residues (Figure 3.5B). Mutagenesis of the

ZnF-C cysteines to alanines in Acme53ZnF-C4C→4A reduced BV production to the same level as ZnF-C deletion at 0, 6, and 12 hpt with undetectable titres of less than 1

87

1E+05 1E+05 Ac+ Ac+ △ME53 △ME53 1E+04 1E+04 △ZnF-N △ZnF-C ZnF-N C-->A * ZnF-C C-->A * 1E+03 1E+03

1E+02 1E+02

1E+01 1E+01

1E+00 1E+00 0 6 12 18 24 0 6 12 18 24

Figure 3.5: The effect of ME53 zinc coordination on BV production. Viral growth curves comparing deletion of ZnF domains with the mutation of the putative ligand binding residues. (A) The effect of mutating the conserved cysteine residues of the ZnF-N in BV production (open triangle) compared to deletion of the whole Zn2+ coordinating sequence from residues 170-209 (open square). The BV growth curves are compared to deletion of the entire me53 (open diamond) and WT (black square). (B) Comparison of ZnF-C cysteine mutagenesis on BV production (black circle) and ZnF-C deletion (x markers) to full me53 deletion (open diamond) and WT (black aquare) BV production. Significance is indicated by a black asterisk. The data points above the asterisk are significantly different (p<0.05, student’s two-tailed paired t-test) than those below.

88

TCID50/mL for all three time points respectively. However, by 18 hpt, Acme53ZnF-C BV

2 titre increased by 140-fold to 1.42 x 10 TCID50 units/mLbut the Acme53ZnF-C4C→4A

BV titre was a significant 111-fold lower at 1.23 TCID50/ml. At 24 hpt, mutagenesis of the

1 cysteine residues resulted in a titre of 7.62 x 10 TCID50/mLwhich was 223-fold lower than

4 the Acme53ZnF-C titre of 1.7 x 10 TCID50/ml. There was a significant difference

(p<0.05) in BV titre at 18 and 24 hpt when the ZnF-C cysteine residues were mutated compared to whole domain deletion, indicating that zinc coordination by the domain is critical for ME53 function.

3.3.5 Selection of viral genes for gene expression analysis

ME53 contains two zinc binding domains and translocates to the nucleus at late times post infection (12 – 18 hpi) via the NTS, suggesting a transcriptional role during virus infection. Since ME53 does not access the nucleus until late times post infection, its direct influence on transcription, if any, would be detected at late times. Hence, 12 and

18 hpt were the time points chosen for RT-qPCR analysis. The mutated variants of ME53 used for BV production analysis were also used for the analysis of viral gene transcription to further characterize the conserved domains’ roles, if any, in transcription during virus infection.

Several viral genes were selected for analysis based on the promoter consensus sequences identified within 300 bp upstream of the translation start site, promoter sequence similarities, and expression profiles. Prior work with ME53 noted the effect of me53 deletion on expression of 30 viral genes classified according to function. As opposed to examining all 30 viral genes previously selected, the promoter consensus sequences were identified within 300 bp upstream of the translation start site of all 30

89 genes and used to select 6 representative genes for expression analysis. The consensus motifs upstream of the 6 selected genes are presented in Table 3.4 and are discussed in conjunction with their expression profiles.

The immediate early gene ie1 is essential for BV production and is expressed immediately following virus infection until ~36 hpi. Its promoter sequence contained seven identified early consensus motifs recognized by host RNA polymerase and four early/late consensus motifs recognized by the viral RNA polymerase complex. Two components of the viral RNA polymerase complex, lef-8 and lef-9, were selected for analysis as they are both expressed from 12 hpi and have similar functions but the early consensus sequences upstream of the ORF are different. There were eight early consensus motifs identified upstream of lef-8 compared to three early motifs upstream of lef-9. Both viral

RNA polymerase subunit genes had four late consensus motifs likely responsible for their high expression at 36 hpi (Acharya and Gopinathan, 2002). Two essential late genes were also selected for analysis. Vp80 encodes a minor nucleocapsid protein oriented at one end of the nucleocapsid that is important for viral DNA packaging and nucleocapsid transport to the cell periphery (Marek et al., 2011). Seven early consensus motifs were found upstream of the ORF start site compared to one early/late and two late motifs.

The second late gene chosen for analysis was gp64, encoding the major glycoprotein found in only budded virus membranes (Wang et al., 2010). The gp64 promoter region contained 15 consensus motifs in total with 7 motifs classified as early and 8 belonging to the early/late and late categories. The only very late gene being investigated was odv-e25 an essential gene encoding a component of the ODV membrane that is also found in the BV membrane and is expressed at very late times

90

Table 3.4: Tetranucleotide promoter consensus motifs identified upstream of the ORF with the “A” of the ATG start codon representing +1. Early Early/Late Late Expression Gene TATA CAGT CATT CAAT TCGT GTAG AGTC TAGG TAAG TATAAGG Pattern Immediate-early ie-1 -173 -51 -201 -290 -90 -82 -85 -195 -229 -93 -103 -20 Early Lef-8 -264 -100 -230 -150 -277 -16 -173 -246 -163 -205 -114 -94 Early Lef-9 -178 -99 -190 -257 -20 -151 -94 Late Vp80 -14 -275 -218 -72 -188 -6 -213 -106 -36 -21 Late Gp64 -171 -39 -143 -286 -91 -38 -90 -139 -70 -56 -126 -68 -22 -66 -19 Very late Odv-e25 -295 -155 -206 -232 -109 -193 -119 -221 -221 -117 -141 -90 -106 -70 -53 -33

91 post infection. Similar to gp64, 16 promoter consensus motifs were found in the promoter region of odv-e25 with 10 considered early consensus motifs and 6 considered early/late or late motifs. The diversity in gene expression and promoter consensus motifs provided distinctions between the viral genes for more specific analysis of ME53 as a transcription factor.

Baculovirus promoter sequences of homologous genes are more divergent than the gene coding regions. This increased dissimilarity represents the importance of promoter consensus motifs for regulation of gene expression (Xing et al., 2005). In addition, promoter consensus motifs tend to be clustered in regions of 5-20 bp suggestive of transcription factor binding sites. For example, ie1 has a group of five tetranucleotide promoter consensus motifs clustered 80-100 bp upstream of the ATG codon, TATA,

CAGT, TAGG, CATT, and TCGT. It is also suggested that these regions could be important for translation initiation factor, or ribosomal protein binding and ribosome recruitment (Xing et al., 2005). The low sequence similarity between the viral promoter sequences of the genes selected for analysis is shown in the sequence identity matrix in

Table 3.5. The highest sequence identity was between the vp80 and odv-e25 promoters at 46%. The other promoters had between 32 – 44% sequence identity thereby removing the variable of promoter sequence similarity as an influence on gene expression.

3.3.6 Primer amplification efficiency verification

The quantitative analysis of gene expression using RT-qPCR is a robust and accurate method if the appropriate controls are employed. The most common controls are internal controls that provide a reference gene to act as a loading control, for normalization of the template input between samples. Internal controls need to be stably

92

Table 3.5 Nucleotide sequence identity matrix of the promoter sequences 300 bp upstream of the translation start site of viral genes selected for qRT-PCR analysis.

Lef-9 Lef-8 Gp64 Vp80 Ie1 Odv-e25

Lef-9 100.00 43.73 33.46 36.25 32.93 32.51

Lef-8 100.00 36.15 39.44 34.96 37.30

gp64 100.00 41.29 37.50 38.78

vp80 100.00 41.85 45.79

ie1 100.00 43.84

Odv-e25 100.00

93 expressed under varying conditions of stress or treatments and should be expressed at levels similar to the target genes. Common reference genes include -actin and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) but they are unsuitable during baculovirus infection of Sf9 cells because their expression levels change significantly during infection. Host ribosomal RNA 28S, and host gene ecdysoneless, a cell cycle regulator, are more appropriate as reference genes as they maintain constant expression levels throughout AcMNPV infection (Salem et al., 2014). For this reason, 28S rRNA was chosen as the reference gene for qRT-PCR analysis of viral gene transcription.

Another important consideration in qRT-PCR analysis is the primer amplification efficiency of the internal control primers and target gene primers. Primers sets must amplify between 90-110% to ensure accurate normalization and analysis of RT-qPCR data (Taylor et al., 2010). To confirm the amplification efficiency of the primers being used, the difference in CT value of the target gene and reference gene at different concentrations of template should remain the same. A comparison of the viral gene target primers CT values and 28S rRNA CT values at 250, 100 ng, and 50 ng of template (Figure 3.6). Each viral gene selected for qRT-PCR analysis had desired amplification efficiencies as indicated by the standard line correlation coefficients (r2) being over >0.980 for all six viral genes and 28S rRNA indicating that the primer sets described in Table 3.3 were suitable for use in transcriptional analysis (Taylor et al.,

2010).

3.3.7 Transcription analysis

The immediate early gene, ie1, is a general transactivator that is necessary for the initiation of viral infection (Guarino and Summers, 1987). The qRT-PCR data analysis for

94

35 28S ie1 R² = 0.9971 lef-8 R² = 0.9986 lef-9 R² = 0.9991 vp80 30 R² = 0.9862 gp64 R² = 0.9999 odv-e25 R² = 0.9906

25 Ct Ct value

20

R² = 0.9911

15

10 1.5 1.75 2 2.25 2.5 log(cDNA template concentration)

Figure 3.6: Analysis of primer amplification efficiencies for RT-qPCR. Primer amplification efficiency analysis to validate the use of viral gene primers and host 28S rRNA primers for an RT-qPCR assay. Several concentrations of template (cDNA), 2 and the corresponding CT values of each primer set were plotted and the r value for the line of best fit of each primer set was used as a threshold for efficient expression (r2 > 0.980.

95 ie1 in the presence and absence of immediate early gene me53 is showed that at 12 hpt, ie1 transcript levels were not significantly affected by ME53 or any of its domains (Figure

3.7). A similar result is seen at 18 hpt, other than a significant 5.7-fold increase in ie1 transcripts caused by deletion of the L region suggesting a direct repressive function of the L region on the ie1 promoter, or an indirect repressive effect possibly by binding to and inactivating transcription factors. In the absence of the L domain, ie1 transcript levels increased significantly from 12 to 18 hpt, indicating that as ME53 translocated to the nucleus, the L domain has a stronger repressive influence on ie1 gene expression.

Otherwise, ie1 transcript levels were relatively stable from 12 to 18 hpt in the presence and absence of ME53 and its conserved domains, consistent with prior observations.

The first viral RNA polymerase subunit gene analyzed was lef-8 whose promoter contained 8 early consensus motifs and 4 late consensus motifs. Lef-8 transcript levels at 12 hpt were not significantly different in the presence or absence of ME53 and its domains (Figure 3.8). However, at 18 hpt, absence of the L domain resulted in a significant increase in lef-8 transcript levels from a 1.6-fold decrease at 12 hpt compared to WT to a 12.5-fold increase in transcripts compared to WT. This resulted in a significant increase in lef-8 transcript levels at 18 hpt compared to WT. In addition, mutation of the

ZnF-N cysteine residues to alanine, resulted in a significant decrease in lef-8 transcript levels from 12 to 18 hpt.

The second RNA polymerase subunit gene, lef-9, which has three early consensus motifs in its promoter region had similar expression profiles to lef-8 but lower magnitudes of fold change (Figure 3.9). The deletion of the L region resulted in an increase in lef-9 transcript levels from a 1.47-fold decrease relative to WT at12 hpt to a 9.33-fold increase

96

ie1 * 7 * 12 hpt 18 hpt 6

5

4

3 change / construct WT

- 2 Fold 1

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N CA ZnF-C CA Bacmid constructs

Figure 3.7: Relative transcript levels of ie1. Relative transcript levels of immediate early gene ie1 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

97

* lef8 15 * 12 hpt 18 hpt 14

13

12

11

10

9

8

7 change construct / / change WT construct

- 6 Fold 5

4 * 3

2

1

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N ZnF-C CA CA Bacmid constructs Figure 3.8: Relative transcript levels of lef-8. Relative transcript levels of viral RNA polymerase subunit gene lef-8 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

98

lef9 * 12 hpt 18 hpt * 11 *

10

9

8

7

6

5

change construct / / change WT construct -

Fold 4

3

2

1

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N CA ZnF-C CA Bacmid constructs

Figure 3.9: Relative transcript levels of lef-9. Relative transcript levels of viral RNA polymerase subunit gene lef-9 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

99 compared to WT at 18 hpt. This resulted in a significant difference between lef-9 transcript levels at 18 hpt between WT and Acme53L transfected cells. Mutation of the ZnF-N cysteine residues also affected lef-9 transcript levels similar to lef-8 transcript levels with a significant 3.33-fold decrease observed at 12 hpt between Acme53ZnF-N and WT transfected cells. There is a 1.34-fold difference between lef-8 and lef-9 transcript levels at 18 hpt in Acme53L transfected cells. The 12.5-fold increase of lef-8 transcripts versus the 9.3-fold increase in lef-9 transcripts in the absence of the L region may be due to the

8 early consensus motifs upstream of lef-8 compared to the 3 early motifs upstream of lef-9. The early motifs are upstream of genes transcribed by host RNA polymerase and may implicate the L region as a binding domain for host polymerase subunits or cofactors to down-regulate the transcription of early viral genes to encourage transcription of viral late genes.

The late gene vp80, which is essential for nucleocapsid formation, was significantly upregulated, though only by 2.1-fold in Acme53ZnF-N transfected cells at 12 hpt compared to WT (Figure 3.10). At 18 hpt, only the absence of the entire me53 resulted in a significant 3.8-fold decrease in vp80 transcript levels. The only significant change between time points for a construct occurred in Acme53ZnF- C transfected cells with a

2.5-fold decrease in vp80 transcript levels from 12 to 18 hpt. The expression of the second late gene gp64, the major BV envelope glycoprotein, was also significantly affected at 12 and 18 hpt by the ZnF-N domain (Figure 3.11). Gp64 transcripts were increased by 2.2- fold at 12 hpt and 2.1-fold at 18 hpt in the absence of the ZnF-N domain. Interestingly, at

12 hpi, deletion of the entire me53 did not significantly reduce gp64 transcript levels, nor

100

vp80 * 3 * 12 hpt 18 hpt *

2 change / construct WT

- 1 Fold

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N CA ZnF-C CA Bacmid constructs

Figure 3.10: Relative transcript levels of vp80. Relative transcript levels of BV and ODV associated late gene vp80 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

101

gp64 4 * * 12 hpt 18 hpt 3

2

change construct / / change WT construct -

1 Fold

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N CA ZnF-C CA Bacmid constructs Figure 3.11: Relative transcript levels of gp64. Relative transcript levels of BV major envelope glycoprotein gene gp64 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

102

odv-e25 * 12 hpt 18 hpt 6 *

5

4

3 change construct / / change WT construct

- 2

* Fold 1

0 Ac+ △ME53 △NTS △ZnF-N △ZnF-C △L △△ZnF ZnF-N CA ZnF-C CA Bacmid construct

Figure 3.12: Relative transcript levels of odv-e25. Relative transcript levels of ODV envelope gene odv-e25 were measured by RT-qPCR and normalized to internal gene 28S rRNA in Sf9 cells transfected with 9 different bacmids containing variable mutations of immediate early gene me53 at 12 (light grey) and 18 (dark grey) hpt, respectively. The bars represent standard deviation between two biological replicates. Significance between bacmid constructs and time points are indicated by an asterisk (p<0.05, student’s two-tailed paired t-test).

103 did deletion of the L region but every other domain resulted in at least a 2-fold decrease in BV production that was maintained through 18 hpt.

The product of the very late gene included in the experiment, odv-e25, is a component of BV and ODV virions. Deletion of the ME53 ZnF-N domain at 12 hpt resulted in a 4.04-fold increase in odv-e25 transcript levels; however, deletion of both zinc finger domains in Acme53ZnF transfected cells resulted in a 3.73-fold decrease in transcript levels (Figure 3.12). Although not statistically significant compared to WT, the ZnF-C domain deletion resulted in a 2.7-fold decrease in transcript levels possibly driving the significant reduction in odv-e25 gene expression at 12 hpt in Acme53ZnF transfected cells. At 18 hpt, Acme53ZnF-N transfected cells still had significantly elevated levels of odv-e25 transcripts. The contrary effects seen with the zinc finger domain deletions on odv-e25 transcript levels may indicate that the zinc finger domains have discrete functions and the by-products from their involvement in other viral infection pathways like budding or protein-protein interactions, may be indirectly antagonistic with respect to transcription.

Mutagenesis of the cysteine residues of the ZnF-N did not mimic the increase in transcript levels seen with full ZnF-N deletion suggesting that the zinc finger structure is not the apparent cause of gene repression, and the continuity of repression from 12 to 18 hpt may indicate instability of ME53 resulting from the deletion of 39 aa within the protein.

Discussion

AcMNPV me53 is a conserved gene found in all alpha and betabaculoviruses with several domains that are universally conserved between both genera. Me53 is an immediate early gene expressed immediately following infection until ~36 hpi via a dual early/late promoter motif and is essential for optimal BV production (Knebel Morsdorf et

104 al., 1996; deJong et al., 2009). During early times post infection, ME53 is predominantly cytoplasmic but begins to translocate to the nucleus at 12 hpi corresponding with the onset of viral DNA replication, and subsequent late gene expression (Liu et al., 2016).

During late times post-infection, 18+ hpi, ME53 is localized in the nucleus and at distinct plasma membrane foci (de Jong et al., 2011). The membrane localization is dependent on the BV envelope glycoprotein GP64, and the foci are suspected budding sites (de

Jong et al., 2011). ME53 contains several domains that are crucial for overall BV production and are intrinsic to its function, including the nuclear translocation sequence and the conserved L region from aa 278-302 (Liu et al., 2016; de Jong, 2011). Deletion of the C-terminal zinc finger of ME53 reduces BV production by 50% at 72 hpi but has not been investigated in terms of the initial viral infection cycle (de Jong, 2011). Similarly, the

ZnF-N domain was only recently identified, and the full domain has not been characterized whatsoever. The coordination of zinc by zinc finger domains is fundamental to their function, and the significance of probable metal ligand-binding cysteine residues on ME53 function and BV production remains unknown. Here, I dissect the role of ME53 domains and their conserved cysteine residues during the first 24 hpt. In addition, the localization of ME53 to the nucleus and the presence of two C4 zinc finger domains is suggestive of a transcriptional role during infection. Therefore, I also investigated the role of ME53 and its domains in the transcription of several essential viral genes.

The evolutionary analysis of ME53 showed the discrete separation of ME53 into three clades representing the Group I alphabaculoviruses, the Group II alphabaculoviruses, and betabaculoviruses. The division of ME53 between Group I and

Group II alphabaculoviruses indicates that ME53 predates the evolutionary divergence of

105

Group I alphabaculoviruses from Group II alphabaculoviruses that resulted in the addition of 11 open reading frames to the Group I alphabaculoviruses (Miele et al., 2011). Several hypotheses for baculovirus evolution exist. The first hypothesis is that baculoviruses originated in Lepidopteran hosts and were randomly horizontally transmitted to other insect orders. However, baculoviruses group very well with their host orders with little variations using several different gene products including DNA Pol, polyhedrin, and cathepsin (Herniou et al., 2004). The second hypothesis claims that baculoviruses originated and evolved with arthropods leading to cocladogenesis of the viruses and their hosts (Herniou et al., 2004). Of these, the second is more plausible as baculoviruses group with their host orders, but the proteins used for phylogenetic analysis do not always root to the more ancient Hymenopteran order. The third hypothesis is a hybrid of the first two hypotheses and postulates that baculoviruses were initially transmitted between hosts of several orders and host-dependent evolution resulted in speciation (Herniou et al., 2004). Therefore, phylogenies result in order-specific groupings but do not necessarily reflect the evolutionary pattern of the hosts. Both alpha and betabaculoviruses infect

Lepidopteran hosts suggesting that ME53 coevolved with Lepidoptera. A major difference in ME53s between the Group I and Group II alphabaculoviruses is that the Group II viruses lack the first 80 amino acids present in Group I. Similarly, the betabaculovirus

ME53 lack the first 110 amino acids present in Group II alphabaculoviruses (Liu et al.,

2016). Since ME53 has no homologues in baculoviruses infecting insects of other orders, or outside of the Insecta class, it is difficult to root the ME53 phylogenetic tree to determine if the 80 to 110 amino acids at the N-terminus of ME53 in Group I alphabaculoviruses is a result of an addition, or if their absence is the result of a deletion from a Group I ME53.

106

Also, granulovirus ME35 has not been investigated whatsoever, and its function may be significantly different from that of AcMNPV ME53. It would be interesting to determine if a betabaculovirus me53 homologue could rescue the Acme53 phenotype as the first

100 amino acids of the N-terminus are not essential for AcMNPV BV production in cell culture (de Jong, 2011; Liu et al., 2016).

The conservation of ME53 and several of its domains in all alpha and betabaculoviruses suggested important functional roles during infection, and was the rationale for domain deletion or mutation in BV production analysis. Deletion of me53, as well as the NTS, ZnF-N or L domains separately, significantly reduced BV production by at least 99.6% compared to WT. Results of deletion of the entire me53 was consistent with the previously reported ~10,000-fold reduction in BV output and demonstrates the essential function of ME53 during infection for optimal BV production (de Jong et al.,

2009). The similar reduction in BV production between the NTS deletion and knock out implies that the nuclear localization of ME53, and therefore its role in the nucleus, is a dominant function of the protein. The NTS might act solely as a localization signal or may be multifunctional in the nucleus. For example, the picornavirus, foot-and-mouth disease virus RNA-dependent-RNA-polymerase contains a non-canonical NLS at the N-terminus that is required for nuclear import as well as specific nucleotide recognition for incorporation during template replication (Sanchez-aparicio et al., 2013; de la Higuera et al., 2018). The ZnF-N and L domains are essential for ME53 function, as demonstrated by the significant affect their deletion had on BV production. Their role during viral infection is uncharacterized, but they may act as binding domains for host or viral proteins

107 for BV assembly based on the binding predictions from Chapter 2, or they may be involved in the regulation of viral gene expression.

Interestingly, deletion of the ZnF-C domain results in a delay in BV production between 12-24 hpt, but BV titre recovers at 24 hpt to 50% of WT. This delay may indicate a cytoplasmic role for the ZnF-C since BV production increases with translocation of

ME53 to the nucleus. However, although BV titre recovers to 50% of WT at 24 hpt, it remains at only 50% of WT through 72 hpt suggesting a more generalized role during infection perhaps with a focus in the cytoplasm during early times post-transfection.

Herpes simplex virus tegument protein VP16 serves several functions during infection with domains specific to each function. One function is the impairment of the transcriptional activation at the C-terminus of VP16 which results in low levels of immediate early gene expression and the establishment of latency but does not impair virus propagation in tissue culture via the lytic cycle (Tal-Singer et al., 1999; Mossman et al., 2000). VP16 is initially diffuse throughout the cell but localizes to the nucleus around

6 hpi with punctate localization in the cytoplasm visualized around 8 to 10 hpi. Similarly,

ME53 ZnF-C function may not be required for AcMNPV virulence in cell culture but may remain important for in vivo infection.

The two ME53 zinc finger domains are separated by 170 amino acids suggesting discrete functions for each, but the tertiary structure of ME53 may bring them within proximity for complementary or shared functions. A complementary function is supported by the negative synergistic effect on BV production observed with deletion of both zinc finger domains from 12 – 24 hpt. Deletion of both zinc fingers reduced BV output by 7.4- fold compared to the ZnF-N only deletion, and 960-fold compared to the ZnF-C only

108 deletion at 24 hpt. The synergistic effect may be due to the involvement of both zinc fingers in the same function or could be a result of the impairment of two separate functions carried out by each zinc finger domain separately. Retroviral nucleocapsid proteins commonly contain two zinc finger domains essential for RNA packaging in the nucleocapsid (South et al., 1990). Similar to ME53, the N-terminal zinc finger position of the HIV-1 nucleocapsid protein is more sensitive to alterations with respect to replication than mutation of the C-terminal zinc finger domain (Gorelick et al., 1996). The HIV-1 nucleocapsid protein zinc finger domains are not biologically equivalent, but both are required for efficient viral replication and are considered complementary to each other’s functions (Guo et al., 2000). Specifically, mutagenesis of the residues involved in zinc binding in the HIV-1 nucleocapsid protein render virions replication deficient, non- infectious and inhibit genome packaging (Goerlick et al., 1990; Goerlick et al., 1996; Guo et al., 2000). Similarly, inhibition of zinc coordination in the ZnF-N zinc finger of ME53 by cysteine mutatgenesis limited AcMNPV BV production to a level similar to that of deletion of the entire domain. However, alanine mutagenesis of the ZnF-C zinc ligands caused a significant reduction in BV production compared to deletion of the ZnF-C domain. This might be due to the necessary coordination of zinc by cysteine residues to stabilize the tertiary structure of ME53, while deletion of the entire domains alters the structure less severely than mutation of structurally important residues. Additionally, amino acids upstream or downstream of the cysteine residues may require the ZnF-C structure for their interaction with putative binding partners, and the absence of structure may cause transient interactions to become permanent or vice versa.

109

ME53 is a component of AcMNPV BV nucleocapsids, and its localization to membrane foci suggests a possible role in BV production as a matrix protein by bridging the nucleocapsid and envelope of BVs for virion egress. However, the reduced levels of infectious BV that are produced in the absence of me53 indicate that ME53 does not play an essential role in BV production. This is similar to the baculovirus core protein AC92, which is not essential for BV production but its deletion results in a low level of infectious

BV (Nie et al., 2011). AC92 is a component of ODVs and has flavin adenine dinucleotide linked sulfhydryl oxidase activity, which is presumed to control the redox state of binding partners and formation of disulfide bonds to encourage abnormal activity of host enzymes during infection (Deng et al., 2007; Long et al., 2009; Nie et al., 2011). AC92 is detected in Sf9 cells from 18 to 72 hpi and initially localizes in the cytoplasm but translocates to the nucleus around 24 hpi and concentrates in the ring zone, a dense region around the periphery of the nucleus where transcription, DNA replication, and nucleocapsid assembly take place (Passarelli and Guarino, 2007; Carstens, 2009). Research recently completed in our lab by Dr. Emine Ozsahin identified ME53 in the ring zone of AcMNPV infected cells. The presence of ME53 in the nucleus before, and during, the onset of viral

DNA replication at ~12 hpi, as well as the important function of two zinc finger domains during BV production strongly supports a putative role for ME53 in the transcription of viral genes required for nucleocapsid assembly and BV egress.

Six viral genes were selected for transcriptional analysis via RT-qPCR at 12 and 18 hpt, and include immediate early gene ie1, viral RNA-polymerase subunit genes, lef-8 and lef-9, BV component genes vp80, and gp64, and the BV and ODV associated protein gene odv-e25. Several domains of ME53 had significant effects on viral gene expression

110 including the L region, ZnF-N domain, and ZnF-C domain. Interestingly, deletion of the entire me53 significantly affected only vp80 expression at 18 hpt, and deletion of the NTS did not significantly affect the expression of any viral genes. No domain deletions of ME53 mimicked the downregulation of vp80 with complete me53 deletion, particularly the NTS, suggesting that the mechanism of downregulation of vp80 is not nuclear. Although the zinc finger domains affected expression of some genes but not others, and not synergistically, suggests an indirect influence on transcription by an unknown mechanism. However, deletion, or mutation, of several ME53 conserved domains, resulted in an increase or decrease of other viral gene transcript levels supporting a multifunctional protein model.

The transcript levels of immediate early gene ie1 were negatively affected by the L region of ME53, with a 6-fold increase in ie1 transcripts observed at 18 hpt when the L region was deleted. Similarly, deletion of the L region increased lef-8 and lef-9 transcript levels at 18 hpt by 12.5-fold and 9.3-fold, respectively. The L region may, therefore, act as a direct repressor of transcription at 18 hpt or, may bind and sequester a transcription factor for the down-regulation of gene expression. In reverse genetic analyses, IE2, IE1, and IE0, positively regulate me53 transcript levels, while ME53 is not directly associated with the regulation of other immediate early genes (Ono et al., 2015). This could indicate that ME53 acts as an intermediate between immediate early gene products IE1 or IE2, and transcriptional regulation of viral genes via the L domain to inhibit overexpression of apoptosis-inducing viral genes such as ie1 (Prikhod’ko and Miller, 1999). For example,

IE2 is negatively autoregulated via auto-ubiquitylation mediated by a RING finger domain, and IE2 accumulation results in the inhibition of BmNPV infection exemplifying the

111 importance of negative regulation in baculovirus infection (Imai et al., 2005). Interestingly, the L domain affected the expression of genes with predominantly early gene consensus motifs and at most one late consensus motif. Host transcription factors presumably recognize the early consensus motifs for viral gene expression prior to the expression and assembly of the viral RNA polymerase complex, therefore implicating the L domain of ME53 in the direct or indirect regulation of viral genes under the control of early promoter consensus motifs.

Lef-8 and lef-9 transcript levels were affected at 12 hpt by alanine mutagenesis of the ZnF-N domain cysteine residues. Putative inhibition of zinc coordination, via mutagenesis of cysteine codons, resulted in a 3-fold decrease in lef-8 transcript levels at

12 hpt compared to 18 hpt, and 3.3-fold decrease in lef-9 transcript levels at 12 hpt compared to WT and implicates the ZnF-N domain in the positive genetic regulation of two RNA polymerase subunits lef-8 and lef-9. Interestingly, deletion of the whole ZnF-N domain resulted in the upregulation of vp80 at 12 hpt, gp64 at 12 and 18 hpt, and odv- e25 at 18 hpt, respectively. Alanine mutagenesis of the ZnF-N did not mimic the effect of

ZnF-N deletion on vp80, gp64, or odv-e25 expression, indicating that residues within the

ZnF-N have functional roles that need to be further characterized. The upregulation of late genes due to ZnF-N deletion could be indirect due to the downregulation of lef-8 or lef-9 due to ZnF-N cysteine codon mutations, and the requirement of lef-8 and lef-9 for viral late gene expression as components of the viral RNA polymerase complex.

However, the discontinuity in the expression of vp80, gp64, and odv-e25 between the

ZnF-N alanine mutagenesis and ZnF-N deletion is inconsistent with this observation and requires further investigation.

112

Several conclusions regarding the functional relationships between ME53 domains can be drawn. First, the zinc finger domains are not involved in the synergistic regulation of any viral genes analyzed in this study and were occasionally antagonistic. Therefore, the synergistic effect of their deletions on BV production is likely due to biologically separate functions that are discretely important for BV production. In addition, the cysteine to alanine mutagenesis for each zinc finger domain, which putatively inhibits the coordination of zinc and therefore the structure of these domains, did not mirror the effect of full domain deletion on viral gene expression thereby signifying the importance of zinc- finger associated residues, other than the zinc ligands, in ME53 function. Finally, the effect of the ZnF-N and L domains on viral gene expression is likely not due to the nuclear localization of ME53, since impairment of nuclear translocation by NTS deletion did not replicate the effects of ZnF-N and L deletions, nor did NTS deletion have any significant effect on viral gene expression whatsoever. Therefore, the nuclear localization of ME53 appears to be exclusively important for mechanisms such as BV assembly via genome packaging, nucleocapsid assembly, nuclear egress and trafficking to the membrane for cellular egress, as supported by the localization of ME53 to plasma membrane foci in a

GP64-dependent manner (de Jong et al., 2011). Also, ME53 zinc finger domains have discrete functions during viral infection, with the coordination of zinc and therefore the structure of the ZnF-C domain functionally more critical for BV production than conserved residues within the domain. The ZnF-N and L domains are recently identified important functional domains and support a multifunctional role for ME53 during AcMNPV infection.

113

Chapter 4: ME53 associates with the ribosome in virus infected cells although RACK1

presence at the ribosome is depleted

Introduction

Translational inhibition is a common anti-viral defense mechanism employed across several kingdoms (McCormick and Khaperskyy, 2017). Baculoviruses are known for their high levels of protein expression during late times post infection and therefore must either actively participate in translation via a baculovirus protein(s) or constrain signaling pathways that lead to translational inhibition. The archetypal method of translational shutoff involves protein kinase R (PKR)-mediated phosphorylation of eukaryotic initiation factor 2 (eIF2). PKR is one of four eIF2 kinases found in vertebrates that recognizes dsRNA, a frequent intermediate of viral replication, and phosphorylates eIF2 to prevent the formation of the eif2-GTP ternary complex that provides the initiator methionine tRNA to the pre-initiation complex (McCormick and Khaperskyy, 2017). The baculovirus protein PK2 is a truncated homologue of eIF2 kinases that interacts with PKR in vitro (Li et al., 2015). Interestingly, insects do not encode homologues of PKR, and instead have homologues for PKR-like endoplasmic reticulum kinases (PERK), general control nonrepressible-2-kinases (GCN2s), and heme-regulated inhibitor kinases (HRI) (Li et al.,

2015). In phylogenetic analyses, PK2 groups with lepidopteran HRI-like kinases and likely originated from an insect HRI-like kinase during a horizontal gene transfer event. EIF2 kinases require homodimerization via the N-terminal lobe for self-phosphorylation prior to eIF2 phosphorylation. The larger C-lobe of the kinases recognize eIF2 and other substrates and is mimicked by baculovirus protein PK2 (Li et al., 2015). Instead of potentially interacting with an eIF2 kinase to mimic and inhibit dimerization, PK2 binds

114 to eIF2 in baculovirus infected cells to directly inhibit the phosphorylation of eIF2 and prevent its inactivation (Li et al., 2015). PK2 knockout in BmNPV does not result in a lethal phenotype in larval assays but was associated with a 12 h delay for larval death, and an increase in eIF2 phosphorylation compared to WT larval and cell culture infections, respectively. In addition, PK2 knockout did not significantly impact BV production until 72 hpi, indicating that either translational arrest is not an efficient anti-viral mechanism, or baculoviruses have other means of counteracting translational inhibition (Li et al., 2015).

Also, the necessary but not sufficient participation of PK2 in translational regulation suggests that other proteins may also interact with translational machinery and may prevent the inhibition of translation like PK2 or encourage the selective translation of viral transcripts.

The host protein RACK1 is an integral part of the 40S subunit and translation initiation, acts as a signalling hub at the ribosome, and is essential for the production of some viruses including Hepatitis C virus (Nilsson et al., 2004; Majzoub et al., 2014).

RACK1 associates with the 40S ribosomal subunit prior to the formation of the 80S complex and coordinates signaling between PKC, MAPKs, and translation initiation factors for the activation or repression of translation via the phosphorylation of initiation factors. MAPK p38b is directly implicated in PKC-independent translational inhibition via

RACK1 dissociation from the ribosome during proteotoxic stress (Belozerov et al., 2014).

However, inhibition of p38b during BmNPV infection does not affect BV production, suggesting that the p38b translational inhibition pathway in baculovirus-infected cells is already inhibited during virus infection. Interestingly, RACK1 is a putative binding partner of ME53 and implicates ME53 in translational regulation during virus infection, possibly

115 by inhibiting p38b dissociation of RACK1 from the ribosome to prevent translational inhibition during early times post infection. The following preliminary study characterizes

ME53 as a translational regulator in AcMNPV infected cells.

Methods

4.2.1 Ac+GFP:ME53 recombinant bacmid construction

The recombinant bacmid Ac+ME53:GFP, containing ME53 with a C-terminal GFP tag used in Chapter 3, was unsuitable for this project as the me53 open reading frame contains a promoter sequence near its 3’-end followed by a start codon that results in expression of GFP alone. Consequently, for the purpose of using the GFP tag to identify

ME53, an N-terminal GFP:ME53 fusion bacmid was constructed. The recombinant bacmid was assembled using the me53 bacmid as the backbone and the pFACT plasmid as the donor plasmid. The gfp ORF was amplified from the pGEM+T-GFP plasmid using primers GFP-FW-XbaI 5’-GCTCTAGAATGGTGAGCAAGGGCGAG-3’

(XbaI site is underlined), and GFP-RV-PstI 5’-

AACTGCAGCTTGTACAGCTTGTACAGCTCGTCCATG-3’ (PstI site is underlined). The fragment was cloned into pBluescriptSV40, described in Chapter 3.2, using XbaI and PstI.

The me53 promoter (proM) was amplified from the commercial bacmid bMon 14272 using primers ME53pro-FW 5’-GAGCTCAGCGTGTGCGCCGGAGCACA-3’ (SacI site is underlined) and ME53pro-RV 5’-TCTAGATGTAACTGTTAGTTAGCACT-3’ (XbaI site is underlined) and subsequently cloned into the pBluescriptGFPSV40 plasmid using SacI and XbaI. The me53 ORF was also amplified from the commercial bacmid genome bMON14272 using the primers ME53-FW-PstI 5’-

AACTGCAGAACCGTTTTTTTCCAGAGAA-3’ (PstI site is underlined) and ME53-RV-

116

EcoRV 5’-CCGGATATCTTAGACATTGTTATTTACAATATTAAT-3’ (the EcoRV site is underlined). After amplification, me53 was subcloned into the pBluescriptproMGFPSV40 plasmid using PstI and EcoRV to generate pBluescriptproMGFP:ME53SV40.

Following me53 cloning into pBluescriptProGFPSV40, the me53 promoter, gfp

ORF, me53 ORF, and SV40 polyadenylation sequence cassette were subcloned into the pFACT donor plasmid using SacI and XhoI generating pFACTproMGFP:ME53SV40

(Figure 4.1). For recombinant bacmid construction, the donor plasmid and the Acme53 bacmid contains attTn7 transposition sites for homologous recombination. The transposition of proMGFP:ME53SV40 into the me53 knockout bacmid was completed in competent DH10 E. coli cells containing the Acme53 bacmid and the helper plasmid that bears the transposase to facilitate homologous recombination. Approximately 200 –

500 ng of the donor plasmid was incubated with the DH10 competent cells on ice for 30 minutes. The cells were heat shocked at 42°C for 45 seconds and recovered in LB medium for four hours at 37°C before being plated on LB agar plates containing kanamycin, gentamycin, chloramphenicol, tetracycline, IPTG, and X-gal for blue/white colony selection. Kanamycin selects for the bacmid genome, gentamycin selects for the donor plasmid, chloramphenicol ensures that the me53 knockout bacmid is present, tetracycline selects for the transposase-encoding helper plasmid, and IPTG and X-gal ensure that white colonies contain bacmids with an insert that has disrupted the IPTG- induced -lactamase gene. The plasmids and recombinant bacmid constructs were verified by sequencing and GFP expression respectively.

117

4.2.2 Bacmid DNA isolation

Bacmid DNA was isolated as described in section 2.2.7. Briefly, DH10 cells containing the recombinant bacmid were cultured overnight at 37°C with shaking in 3 mLof LB medium containing 50 g/mL Kanamycin, 34 g/mL chloramphenicol, and 7

g/mL gentamycin. The E. coli cells were then pelleted and resuspended in 300 L 15 mM Tris-Cl, 10 mM EDTA, and 100 g/mL of RNaseA. A 0.2 N solution of NaOH and 1%

SDS was used to lyse the cells at room temperature followed by the addition of 300 L of a 3 M potassium acetate solution, pH 5.5, to neutralize the negatively charged backbone of DNA. Precipitated protein and cellular debris were pelleted via centrifugation at 14,000 x g and the supernatant was transferred to a new microcentrifuge tube containing absolute isopropanol to precipitate bacmid DNA. Further centrifugation pelleted the DNA precipitate which was then washed with 70% ethanol to remove any remaining salts. The

DNA pellet was air dried before being resuspended in mqH2O and quantified using a

NanoDrop ND-1000.

4.2.3 Virus amplification and titration

The recombinant bacmid was used to generate AcME53-GFP:ME53 virus as described in Section 2.2.8. Briefly, 1.0 x 106 Sf9 cells were transfected with 5 g purified bacmid DNA using Cellfectin II reagent (Thermo Fisher) and the supernatant was collected at 72 hpi. The supernatant containing the recombinant virus, considered the P0 stock, was amplified by infecting 1.0 x 107 Sf9 cells with the P0 stock and collecting the supernatant of the infected cells at 72 hpi, which is designated the P1 stock. The P1 stock was titred as described in Section 3.2.5 using the end-point dilution assay with occlusion body formation used as the marker for infection. The ability of the bacmid transfection to

118 generate viable virus indicates that the addition of a GFP tag to the amino end of ME53 does not affect the functionality of ME53 during virus replication.

4.2.4 Polysome Profiling

To determine if ME53 associated with the translation complex, polysomes of cells infected with Ac+GFP:ME53 were analyzed. Sf9 cells were seeded in 15 cm culture dishes at a density of 5 x 105 cells/mL and allowed to attach for 24 hours at 27°C.

Following the initial adherence period, the cells were either infected with Ac+GFP:ME53 at an MOI of 5 or left uninfected until the infected cells were harvested at either 12 or 20 hpi. At the point of collection, uninfected cells were 70 – 90% confluent to ensure that translation was still active. At harvest, cells were incubated with 0.1 mg/mL cycloheximide

(Acros Organics, Morris Plains, NJ) for 10 minutes at room temperature to immobilize ribosomes at their positions on the transcripts. Following cycloheximide treatment, cells were scraped from the cell culture dishes and centrifuged at 4,000 x g for 5 min to pellet the cells before being washed twice in ice-cold PBS containing 0.1 mg/mL cycloheximide.

The cells were then resuspended in 1.5 mL RNA lysis buffer (15 mM Tris:HCl (pH 7.4),

15 mM MgCl2, 0.3 M NaCl, 1% Triton X-100, 0.1 mg/mL cycloheximide, and 100 units/mL

RNasein) and incubated on ice for 15 min with occasional vortexing to ensure complete lysis. The lysate was centrifuged at 5,000 rpm at 4°C for 10 min and the supernatant was transferred to a new 1.5 mL microcentrifuge tube before being centrifuged again at 14,000 rpm at 4°C for 10 min. The supernatant from the previous centrifugation was transferred to a new 1.5 mL tube and the OD of the lysate at 260 nm was measured using a NanoDrop

ND-1000. The OD values were used to ensure equal loading of the samples onto sucrose density gradients (7-47% wt/vol). Following loading, centrifugation was performed at

119

180,000 x g with an SW-41-Ti Rotor (Beckman Coulter, Fullerton, CA) for 90 min at 4°C.

Ten 1 mL equal fractions were collected from the top of the gradients after centrifugation using the BR-188 Density Gradient Fractionation System (Brandel, Gathersburg, MD).

The absorbance at 254 nm, representing total RNA was continuously monitored throughout fractionation.

4.2.5 Trichloracetic acid (TCA) Precipitation

Analysis of ribosomal and potential viral proteins following polysome profiling requires the precipitation of proteins from the sucrose gradient fractions. TCA was added at a 1:4 ratio (250 L/750 L) of TCA to sample prior to incubation at 4°C for 10 min. The fractions were then centrifuged for 5 min at 14,000 rpm and 4°C. The supernatant was discarded, and the pellet was washed with 200 L cold acetone and centrifuged again at

14,000 rpm for 5 min at 4°C. The supernatant was discarded, and the acetone wash was repeated. The pellet was then dried by placing the microcentrifuge tubes in a 95°C heat block for 40 sec to evaporate any remaining acetone. The pellet was then resuspended in 40 L SDS loading buffer before being analyzed via SDS-PAGE and Western blot.

4.2.6 SDS-PAGE and Western blot analysis

TCA-precipitated protein samples from the polysome profiling fractions were separated in 12% SDS-PAGE gels for 2 h at 100V before being Coomassie blue stained or transferred to a PVDF membrane for 1 h at 30 mA with fresh buffer (~100 V). The

PageRuler pre-stained protein ladder (Thermo Fisher) was used to determine the molecular weight of proteins in Coomassie blue or Western immunoblot bands. For

Coomassie blue staining, gels were stained with 0.2% Coomassie Blue R-250 (Fisher) in

10% acetic acid, 50% methanol, and 40% dH2O for 30 minutes. The gels were destained

120

in 50°C dH2O for 15 minutes 4 times with shaking. For Western blotting, following protein transfer to the PVDF membranes, the membranes were incubated in blocking buffer consisting of 2% skim milk powder in TBS overnight at 4°C. After blocking, the membranes were rinsed in TBS-T for 10 minutes three times. The membranes were then incubated with primary monoclonal rat anti-GFP (Chromotek, 3H9) for GFP:ME53 detection, or primary monoclonal rabbit anti-RACK1 (Cell Signaling Technology, D59D5), for RACK1 detection, at 1:1000 dilutions in TBS-T with 0.2% skim milk powder. Primary antibody incubation with the membranes was carried out at 20°C with gentle shaking for

1 hour. The membrane was then washed in TBS-T for 5 minutes 4 times before being incubated with the appropriate HRP-conjugated secondary antibodies, anti-rat IgG

(Invitrogen) and anti-rabbit IgG (GE Healthcare, NA934V), respectively. Secondary antibody incubation occurred over 1 h at room temperature with shaking and was followed by washing the blot with TBS-T for 10 min 4 times. Super signal West Pico chemiluminescent substrate (Thermo Fisher) was applied to the membrane to detect the

HRP-conjugated secondary antibodies for the Western blot. The ChemiDoc MP System

(BioRad, Mississauga, ON) was used to detect the fluorescence and visualize the bands.

Results

4.3.1 GFP tagged recombinant bacmid construction and virus amplification

The GFP-tagged ME53 recombinant bacmid was successfully generated as shown by sequencing (Figure 4.1). Sf9 cells were successfully transfected with bacmid

DNA and two rounds of virus amplification resulted in a P1 stock with a BV titre similar to a wildtype construct. The wildtype virus, Ac+proMGFP had an average titre of 2.67 x 107

+ 7 TCID50/mL, and the Ac GFP:ME53 virus titre was 1.85 x 10 TCID50/mL (Figure 4.2). The

121

A

B

Figure 4.1 Construction of GFP:ME53 bacmid (A) Schematic of GFP tagged ME53 recombinant bacmid using the Acme53 bacmid. The gfp ORF (purple) was fused to the N-terminus of the me53 ORF (grey), and both are under the control of the me53 promoter. (B) The gfp only schematic shows gfp under the control of the me53 promoter with an SV40 polyA sequence (green) at the C- terminus

122

5.E+07

4.E+07

3.E+07

2.E+07

1.E+07

0.E+00 Ac+GFP Ac+GFP:ME53

Figure 4.2 Titre of GFP-only and GFP-tagged ME53 viruses using the end point dilution assay method. The Ac+GFP virus titre is not significantly different over two biological replicates (p-value = 0.5, student’s two-tailed paired t-test) from the Ac+GFP:ME53 virus titre indicating that the addition of a GFP N-terminal tag to ME53 does not affect BV production.

123 similar titres indicate that the N-terminal GFP tag on ME53 does not significantly impact the function of ME53 e.g. by disrupting its structure, and that it is appropriate to use for

ME53 detection in polysome profiling.

4.3.2 AcMNPV protein ME53 associates with ribosomes in baculovirus infected cells

The interaction of ME53 with the ribosomal protein RACK1 suggested a role for

ME53 at the ribosome. To determine if ME53 does associate with the ribosome in infected cells, polysome profiling was used to separate ribosomal subunits, monosomes, and polysomes for protein analysis within each fraction of the sucrose gradient. The translational activity of virus-infected cells was compared to uninfected cells that were actively growing and therefore would be translationally active. An investigation of the translational landscape indicated that infection did not cause a decrease in overall translation levels according to polysome profiles, but ME53 was likely associated with the ribosome in AcMNPV infected cells (Figure 4.3).

The polysome profiles at early times post infection (12 hpi) showed that the 40S and 60S subunit peaks had a greater amplitude than the same peaks in uninfected cells, possibly indicating an increase in ribosome subunit biogenesis (Figure 4.3A). The peaks corresponding with the monosomes (80S), light polysomes (LP, 2-4 ribosomes per transcript), and heavy polysomes (HP, 5-8 ribosomes per transcript) were fairly similar in uninfected and early infected cells, although the infected cells showed higher polysome peaks and an additional peak in the HP fraction indicating increased translational activity.

As hypothesized, based on its preliminary interaction with ribosomal protein

RACK1, ME53 was detected in all fractions associated with the subunits, monosomes,

LPs, and HPs during early times post infection. To ensure that adequate protein was

124 present in each sample, the fractions corresponding to the polysomes were pooled. ME53 was located in the soluble fraction from the sucrose gradients, where proteins can be associating with the 40S subunit but are also part of the soluble protein. This could be seen with GFP, used as a negative control to ensure that GFP alone was not interacting with the monosomes or polysomes. GFP was found in only the first two fractions correlating with the 40S and 60S subunits. The presence of GFP:ME53 in the 80S, LP, and HP fractions suggests that it is ME53 only that associates with the ribosome at early times post infection, and not through the GFP tag.

Polysome profiles of infected and uninfected cells at 20 hpi, showed that translation is increased in Ac+GFP infected cells compared to uninfected cells harvested at the same time (Figure 4.4B). Depleted amounts of ribosomal subunits, as shown by the lower peaks associated with the 40S and 60S fractions, and the increase in peak amplitudes of the 80S, and LP fractions at 20 hpi indicated increased translation in infected cells. At 20 hpi, the infected cells also appeared to have more heavy polysomes than uninfected cells with nine peaks visible after the monosome during AcMNPV infection compared to eight polysomes in uninfected cells. The additional heavy polysome indicated that virus-infected cells contained transcripts associated with up to at least 10 ribosomes versus the 9 observed in uninfected cells. Similar to what was seen at 12 hpi,

ME53 was associated with the monosome and light polysome fractions at 20 hpi (Figure

4.3B). ME53 was principally located in the soluble fraction, correlating with its predominantly nuclear localization at 20 hpi, and a small population of ME53 was associated with the monosome peak and LP peaks. GFP alone had a similar pattern at

20 hpi as 12 hpi but was more concentrated in the 60S fraction at late times due to an as

125

A B

Figure 4.3 Polysome profilies of uninfected and GFP:ME53 virus infected cells at 12 and 20 hpi. (A) Polysome profile of uninfected (above) and Ac+GFP virus infected (below) Sf9 cells at 12 hpi with peaks corresponding to the ribosomal subunits (40S or 60S) or monosomes (80S), LP, and heavy polysomes HP. The corresponding Western blots below are from the sucrose gradient fractions of Ac+GFP:ME53 and Ac+GFP virus infected cells respectively. An anti-GFP primary antibody was used with both Westerns to detect ME53 or GFP. (B) Polysome profiles of uninfected (above) or Ac+GFP infected (below) Sf9 cells with the appropriate peaks labelled. The Western blots below used a primary anti-GFP antibody to detect GFP:ME53 in Ac+GFP:ME53 infected cells or to detect GFP in Ac+GFP infected cells respectively.

126 of yet unknown reason. Regardless, free GFP was not seen in the monosome or polysome fractions, compared to ME53, which supported the observation that ME53 associated with the monosomes and polysomes in virus infected cells.

Further work should include a puromycin control, which causes the dissociation of ribsomes and mRNA, resulting in ribosome-associated proteins accumulating in the unbound ribonucleoprotein (RNP) fraction as opposed to polysome fractions (Pestka,

1971). The addition of puromycin to infected and uninfected cells would ensure that ME53 is sedimenting with ribosomes and not other large complexes like polymerase holoenzymes (Kimura et al., 1999). Other controls should include a positive control such as using an antibody against ribosomal protein RPS12, a 40S subunit protein, in addition to RACK1, to confirm the distribution of ME53 between the ribosomal subunits, monosomes, and polysomes. (Rabl et al., 2011)

4.3.3 Bacluovirus infected cells show decreased concentration of RACK1 at the

ribosome than uninfected cells.

The interaction between ME53 and RACK1, and the localization of ME53 with the ribosome suggested that ME53’s association with the ribosome was facilitated by RACK1.

To determine if RACK1 was the likely candidate for ME53 ribosomal localization, Western blots of polysomes from Ac+GFP virus-infected cells were performed to analyze the total amount of RACK1 present in uninfected versus infected cells. In addition, Western blots of polysome profile fractions were performed to determine the localization of RACK1 at the ribsome during baulovirus infection. First, total RACK1 was measured in uninfected and Ac+GFP infected cells at 12 and 20 hpi respectively. Coomassie blue staining of the total soluble protein from the uninfected and infected samples was compared to ensure

127

Figure 4.4: Total concentration of RACK1 in uninfected and infected cellsm and the localization of RACK1 at the ribosome. (A) Total proteins in uninfected and infected Sf9 cells at 12 hpi and 20 hpi. Coomassie staining was used to normalize the total soluble protein input from each sample prior to Western blotting. Primary anti-RACK1 antibody was used to detect total RACK1 in the soluble protein prior to sucrose centrifugation for polysome profiling and to show the amount of RACK1 detected in each sample (shown below each Coomassie Blue stained gel). (B) Anti-RACK1 Western blot of TCA precipitated polysome profiling fractions in uninfected cells showing the ribonucleoproteins (RNP), 40S subunit (40S), 60S subunit (60S), monosome (80S), LPs, and HPs. RACK1 is shown at the expected 36 kDa size in the Western blots indicated by the arrows. (C) Coomassie stain of polysome fraction proteins from uninfected and Ac+GFP infected cells at 12 and 20 hpi.

128 equal loading for Western blot analysis (Figure 4.4A). The total amount of RACK1 in uninfected and infected cells was fairly similar over the two biological replicates. The uninfected cells appeared to have a higher concentration of RACK1 compared to uninfected cells at both 12 and 20 hpi, although the associated Coomassie stain did show that slightly more protein may have been loaded compared to the infected cells. Overall, there was no obvious difference in the total amount of RACK1 between uninfected cells and Ac+GFP infected cells at 12 or 20 hpi.

While the overall concentration of RACK1 appeared to be unchanged (Figure

4.4A), the sequestration of RACK1 into stress granules during viral infection could influence the total amount of RACK1 between the samples to appear unchanged; however, the amount of RACK1 at the ribosome may be altered. Fractions from polysome profiles were used for Coomassie blue staining and Western blots to normalize the amount of protein and determine the amount of RACK1 present at the polysomes of uninfected and infected Sf9 cells. Figure 4.4B shows the amount of RACK1 localized at the ribosome in uninfected and infected cells at 12 and 20 hpi. The uninfected cells contained a typical distribution of RACK1 from the 40S fraction to the HP fractions with increased density around the 80S fraction. At 12 hpi, the Coomassie stain from Figure

4.4C had a similar pattern to the associated polysome profile where there was more protein present in the 40S, 60S, and 80S ribosomal subunit and monosome fractions, which is where RACK1 was located on the Western blot (Figure 4.4B). RACK1 was detected in the LP fractions, but the signal was faint compared to uninfected cells. The faint signal may have been due to the lower amount of protein present in the polysome fractions of the 12 hpi sample compared to the uninfected and 20 hpi samples. The

129 soluble and subunit fractions at 12 hpi did contain more protein than the corresponding fractions in uninfected cells but still had less RACK1 suggesting that RACK1 may have been partially sequestered or dissassociated from the ribosome at early times post infection. The Coomassie blue stain of the 20 hpi sample showed similar amounts of protein to the uninfected cells particularly in the light polysome fractions. However, the amount of RACK1 detected via Western blot was lower at 20 hpi than in uninfected cells but higher than the amount of RACK1 detected at the ribosome at 12 hpi. This may have been due to the increased translational activity seen at 20 hpi versus 12 hpi but it is interesting to note that the translational landscape at 20 hpi was more active than in uninfected cells although there are lower amounts of RACK1. Further work including additional RACK1 Western blots following polysome profiling, and immunofluorescence of RACK1 to observe its intracellular localization during baculovirus infection would be beneficial in understanding the decreased presence of RACK1 at the ribosome in baculovirus infected cells.

Discussion

To study the translational landscape of baculovirus-infected cells and investigate if

ME53 was interacting with the ribosome, a recombinant bacmid containing ME53 with an

N-terminal GFP-tag was generated. The N-terminal tag did not affect BV production, as determined by end-point dilution assays and the clear appearance of CPEs after 7 days.

The GFP-epitope tag was appropriate for visualizing ME53 in cellular fractions from a functional standpoint and unlike GFP:ME53, GFP only was not found to associate with monosomes or polysomes, verifying its suitability for the identification of ME53 in polysome-containing fractions using Western immunoblot (Brengues et al., 2005).

130

Baculovirus infection results in slightly increased translational activity compared to uninfected, actively growing cells. ME53 associates with sucrose gradient fractions corresponding to the monosomes and polysomes of virus infected cells, more so at 12 hpi than 20 hpi aligning with the timing of cytoplasmic localization of ME53. Although

ME53 was easily identified in polysome fractions, several controls need to be applied to confirm that ME53 is associating with the monosomes and polysomes and not a large ternary complex with similar sedimentation rates. The cycloheximide treatment used in this study inhibits translational elongation and prevents dissociation of the ribosome from mRNA. Trapping the mRNA in the ribosome causes the large complexes to remain stable, whereas large polymerase complexes that potentially sediment at the same rate as monosomes and polysomes would not remain stable throughout the lysis and centrifugation processes (Mollet et al., 2008). However, puromycin, which was not used in this study, is most commonly used as a negative control by causing premature termination of translation, and the dissociation of ribosomes from mRNA. Puromycin prevents ribosomal proteins from sedimenting as monosomes or polysomes but rather as separate RNPs. If ME53 is actively participating in translation as suggested by its localization at the ribosome, and its interaction with RACK1, a puromycin control should result in the dissociation of ME53 from the ribosome and it would predominantly localize to the RNP fraction, and so, the interaction of ME53 at the ribosome requires further characterization.

The interaction of ME53 with RACK1 led to the hypothesis that ME53 localization to the ribosome was mediated by RACK1. The total amount of RACK1 and RACK1 localization at the ribosome during virus infection was investigated. Initially, the interaction

131 of ME53 and RACK1 was surmised to prevent RACK1 dissociation from the ribosome during virus infection, which occurs during proteotoxic stress in Drosophila (Belozerov et al., 2014). However, although the total concentration of RACK1 remained unchanged during AcMNPV infection, the RACK1 population at the ribosome was markedly decreased. RACK1 is essential for translation during infections with several viruses including the cap-independent IRES-mediated translation of Hepatitis C Virus (HCV) and the selective translation of Vaccinia virus (VacV) transcripts (Majzoub et al., 2014; Jha et al., 2017). VacV kinase B1 does not modify the entire RACK1 population during infection, just enough to support VacV protein synthesis (Jha et al., 2017). VacV infection modifies human RACK1 through plant mimetic phosphorylation of several residues in the loop region to selectively translate viral mRNAs with extended poly-A leaders (Walsh, 2017).

The lower concentration of RACK1 at the ribosome in AcMNPV infected cells may be due to inefficient customization of the ribosome for the preferential translation of viral mRNA.

One possible method of RACK1 customization is ME53-directed use of the RACK1 binding partner PKCβII to phosphorylate key residues on RACK1, as ME53 does not have any predicted inherent kinase function. A second possibility is that ME53 directly inhibits p38b-mediated dissociation of RACK1 from the ribosome allowing a sufficient subpopulation of RACK1 to continue translation immediately following viral infection, at least until transcription of host genes are downregulated by an unknown mechanism. For example, HSV encodes a virus host shutoff (vhs) protein expressed throughout infection that accelerates the degradation of host and viral mRNAs and prevent host protein synthesis (Lam et al., 1996). Vhs is mediated by the tegument protein VP16 which binds vhs during late times post infection to limit the degradation of viral mRNA and promote

132 the synthesis of viral proteins (Lam et al., 1996). VP16 impacts virus production by limiting translational arrest via vhs thereby increasing the amount of virion structural proteins for efficient assembly and egress (Mossman et al., 2000). ME53 may have a similar functional mechanism whereby each conserved domain has a separate important role throughout the course of baculovirus infection. The putative presence of ME53 at the ribosome will provide new insights into the regulation of protein synthesis in baculovirus- infected cells and may have important applications in the use of BEVS for recombinant protein production.

133

Chapter 5: General conclusions and future directions

AcMNPV ME53 is an immediate early gene expressed from a dual early/late promoter that facilitates its expression from 0.5 hpi to 36 hpi (Knebel-Mörsdorf et al., 1996; de Jong et al., 2009). ME53 is essential for optimal BV production, with inhibition of its late expression resulting in a larger effect than inhibition of early expression, on BV production suggesting a role in nucleocapsid assembly and egress during late times post infection

(de Jong et al., 2009). That ME53 translocates to the nucleus in an infection-dependent manner prior to, and during, viral DNA replication then proceeds to localize to the viral ring zone of infected cells corroborates this observation (Liu et al., 2016). ME53 does not influence DNA replication, and during late times post infection, ME53 localizes to plasma membrane foci in a GP64-dependent manner, further indicating a role in the structural production of BV (de Jong et al., 2011). The nucleocapsids of BVs contain ME53, and

ODVs may also contain ME53 although reports differ by viral and host species (de Jong et al., 2011; Hou et al., 2012). Initial identification of ME53 revealed a region of the C- terminus presumed to contain a C4 zinc finger domain due to the presence of a highly conserved canonical zinc finger motif (Knebel- Mörsdorf et al., 1993). Although the general role of this domain in BV production has been investigated, its presumed role in transcription remained to be determined (de Jong, 2011). The goal of this thesis was to confirm the presence of a zinc-binding domain in the previously identified putative C- terminal zinc finger domain (ZnF-C), as well as investigate a more recently identified conserved putative zinc finger domain closer to the N-terminus (ZnF-N), and further characterize the role of ME53 and several conserved domains during BV production, transcription, and translation.

134

First, the ability of the ME53 putative zinc finger domains to coordinate a zinc ion, and therefore undergo structural changes in the presence of zinc was analyzed using purified peptides and CD. In addition, secondary structure prediction algorithms JPred and

IUPred2A were used to estimate the secondary structure of the domains for comparison with the CD estimated secondary structure. Purified peptides encompassing the putative zinc finger domains were incubated with or without ZnCl2 and spectra were observed from

190-250 nm for distinct changes in secondary structure due to the presence of zinc. The

CD-generated spectra of the ZnF-N peptide showed a distinct difference with the addition of zinc from a predominantly disordered spectra to one with an increased -strand structure. The increase in -strand structure was consistent with the JPred prediction software and resulted in the preliminary classification of the ZnF-N into the treble clef fold family of zinc finger domains. The ZnF-C domain was not as straightforward as the N- terminal domain. CD analysis of the peptide resulted in an unusual spectrum with large discrepancies between the JPred predicted secondary structure and the secondary structure prediction from the CD spectra. The JPred algorithm predicted a -strand structure for the ZnF-C domain, but the CD spectra data was interpreted as α-helix and

-strands with low confidence. Although the CD secondary structure prediction was inconsistent with the JPred prediction, and had a high NRMSD value, there was a distinct change in the CD spectra with the addition of zinc. The change in secondary structure observed with the addition of zinc indicates that the interaction of the ZnF-C domain with

Zn2+ altered its secondary structure, supporting the presence of a functional zinc finger domain.

135

Other structural prediction software for disorder, binding sites, and redox-sensitive regions, predicted that ME53 is predominantly ordered and contains several binding sites in several conserved domains. The first site was the NTS, a presumed binding site for a nuclear chaperone protein, the second site encompassed the N-terminal region of the

ZnF-N domain, and the third site involved a 50 aa region beginning at the C-terminal region of a conserved region termed the “L” domain from aa 278-302. Absolute confirmation of ZnF-N and ZnF-C as zinc finger domains requires a crystal structure of

ME53. The BEVS system was used to successfully express and purify His-ME53 for such future work and should be utilized for structural classification of ME53 to identify any structural homologues and further characterize the function of ME53 during infection.

Phylogenetic analysis of ME53 supported the previous observation that ME53 is conserved in all alpha- and betabaculoviruses (Liu et al., 2016). The phylogenetic tree also confirmed that the betabaculovirus ME53 lacks 110 residues at the N-terminus of

ME53 that are found in Group I alphabaculovirus and lacks 80 residues at the N-terminus of ME53 from Group II alphabaculoviruses, and those residues are the only predicted disordered region of AcMNPV ME53. It is unclear if these 80-110 N-terminal amino acids are the result of an addition or deletion event, but they are not necessary for BV production during alphabaculovirus infection (Liu et al., 2016). Future experiments should investigate if the shorter GV ME53 can rescue Acme53 virus, to potentially determine if the N terminal 110 aa in AcMNPV were added during evolution to give a functional advantage in alphabaculoviruses, or if they were removed during evolution of the betabaculoviruses because they are non-functional.

136

The strong indication of zinc finger domains in ME53, and several conserved domains that may act as binding sites led to the investigation of these regions to establish if the domains are essential for BV production during the initial infection cycle, and if the zinc finger domains are involved in the same or discrete functions. In addition, since zinc finger domains are commonly implicated as transcription factors, several viral genes were selected for analysis in the presence and absence of these domains. More specifically, several conserved domains of ME53, and the zinc finger domains were analyzed throughout the first 24 hpi and their role in viral gene transcription was also investigated.

Individual deletion of the entire me53 as well as deletions of each domain with the exception of the ZnF-C region resulted in a significant decrease in BV production from 12

– 24 hpt. ZnF-C deletion significantly attenuated BV production from 12 – 18 hpt but increased to 50% of WT by 24 hpt. Interestingly, deletion of both zinc finger domains resulted in a greater attenuation in BV production than deletion of either alone suggesting a complementary function of the ZnFs. Mutagenesis of the zinc finger cysteine codons to alanine affected BV production more than whole domain deletion for both zinc finger domains. However, the mutagenesis of the ZnF-C cysteine residues significantly affected

BV production compared to entire domain deletion suggesting that the coordination of zinc, and therefore the structure of the ZnF-C domain is more important for ZnF-C function than residues within the domain. The NTS, ZnF-N, and L domains are all predicted binding regions of ME53, so future work should include the identification of binding partners for ME53 and the potential role these domains have for ME53 localization at the plasma membrane or in the nucleus. In addition, these domains may be optimal candidates for mutagenesis when binding partners of ME53 are identified.

137

Several viral genes were selected for transcriptional analysis based on the promoter consensus sequences upstream of the ATG codon. Whole me53 deletion and deletion of the NTS downregulated only the late expression of vp80 and none of the other selected genes, respectively. The lack of variation in transcription levels in the presence and absence of ME53 in the nucleus suggests that ME53 does not act directly as a generic transcriptional regulator of all viral genes in virus infected cells. This was supported by the effect of deletion of several ME53 conserved domains, which resulted in an increase or decrease of viral transcript levels suggesting a multifunctional protein model that may indirectly influence viral gene transcript levels. Deletion of the L domain had the largest impact on viral gene expression, and consistently affected viral genes with early consensus motifs, while the ZnF-N domain affected only viral genes with late consensus motifs. Interestingly, the zinc finger domains had opposing effects on gene expression, with deletion of the ZnF-N and ZnF-C domain deletion resulting in upregulation and downregulation of viral genes, respectively. The synergistic effect seen on BV production from deletion of both domains is therefore likely not due to a complementary function of the domains. Future work for the characterization of ME53 will benefit from the analysis of total BV production to determine if different ME53 domain deletions result in the release of non-infectious BV, as occurs with the mutagenesis of HIV-1 nucleocapsid protein zinc finger domains (Gorelick et al., 1999).

The final objective of this thesis was the preliminary investigation of ME53 at the ribosome based on yeast-2-hybrid results indicating that the 40S ribosomal subunit protein RACK1 could be a binding partner of ME53. ME53 appears to localize to the ribosome during viral infection based on Western immunoblots of polysome profiles from

138

Ac+GFP:ME53 infected cells, and it was hypothesized that ME53 may be interacting with

RACK1 to prevent its dissociation and sequestration into stress granules. However,

RACK1 localization at the ribosome decreases in infected cells compared to uninfected cells, but the total amount of RACK1 remains the same. This suggests that only a smaller population of RACK1 is required for efficient translation, similar to that in VacV infected cells. ME53 may be associated with the inhibition of RACK1 sequestration or may associate with the ribosome separately for an unknown mechanism. Further work to confirm the ME53 RACK1 interaction using pulldowns, and confirmation of ME53 interaction with the ribosome using the appropriate controls is recommended, as ME53 is the first baculovirus protein implicated to directly interact with the ribosome during baculovirus infection.

The results from this multitier study strongly support a multifunctional model for

ME53, with an immediate early C4 zinc finger protein and several conserved domains responsible for discrete and temporal roles. The NTS may have secondary roles other than nuclear translocation, and the ZnF-N is predicted to be a binding domain, perhaps to late expression factors, or at the plasma membrane. The L region is important for early gene transcription, likely via an indirect mechanism, but further work needs to identify the precise function of this domain. Further work should include the structural analysis of

ME53 using the developed BEVS for confirmation of zinc binding to the zinc finger domains, as well as the investigation of ME53 at the ring zone during viral infection, possibly indicating a role in DNA packaging, as well as at plasma membrane foci as a putative scaffolding protein facilitating membrane attachment and budding.

139

REFERENCES Acharya, A. and Gopinathan, K.P. 2002. Characterization of late gene expression factors lef-9 and lef-8 from Bombyx mori nucleopolyhedrovirus. J. Gen. Virol. 83:2015-2023. Adams, D.R., Ron, D., and Kiely, P.A. 2011. RACK1, a multifaceted scaffolding protein: structure and function. Cell Commun. Signal.9:22. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410. Arimoto, L., Fukuda, H., Imajoh-Ohmi, S., Saito, H., and Takekawa, M. 2008. Formation of stress granules inhibits apoptosis by suppressing stress-responsive MAPK pathways. Nat. Cell Biol. 10:1324-1332. Au, S., Wu, W., and Panté. 2013. Baculovirus nuclear import: open, nuclear pore complex (NPC) sesame. Viruses. 5:1885-1900. Ayres, M. D., Howard, S. C., Kuzio, J., Lopez-Ferber, and M., Possee, R. D., 1994. The complete DNA sequence of Autographa californica nuclear polyhedrosis virus. Virology 202:586-605. Balan, N., Osborn, K., and Sinclair, A.J. 2016. Repression of CIITA by the Epstein-Barr virus transcription factor Zta is independent of its dimerization and DNA binding. J. Gen. Virol. 97:725-732. Belozerov, V., Ratkovic, S., McNeill, H., Hilliker, A., and McDermott, J.C. 2014. In vivo interaction proteomics reveal a novel p38 mitogen-activated protein kinase/RACK1 pathway regulating proteostasis in Drosophila muscle. Mol. Cell. Biol. 34:474-484. Bleckmann, M., Fritz, M.H.Y., Bhuju, S., Jarek, M., Schurig, M., Geffers, R., Benes, V., Besir, H., and ven den Heuvel, J. 2015. Genomic analysis and isolation of RNA polymerase II dependent promoters from Spodoptera frugiperda. PLoS One. E0132898. Blissard G. W. (1996). Baculovirus—insect cell interactions. Cytotechnology. 20:73-93. Blissard, G. W., and Rohrmann, G. F., 1990. Baculovirus diversity and molecular biology. Annu. Rev. Entomol. 35:127-155. Brengues, M., Teixeira, D., and Parker, Roy. 2005. Movement of eukaryotic mRNAs between polysomes and cytoplasmic processing bodies. Science. 310:486-489. Broadley, H.J., Boucher, M., Burand, J.P., and Elkinton, J.S. 2016. The phylogenetic relationship and cross-infection of nucleopolyhedroviruses between the invasive winter moth (Operophtera brumata) and its native congener, Bruce spanworm (O. bruceata). J. Invertebr. Pathol. 143:61-68. Cao, X., Hu, X., Zhang, X., Gao, S., Ding, C., Feng, Y., and Bao, W. 2017. Identification of metal ion binding sites based on amino acid sequences. PLoS One. E0183756 Carstens, E.B. 2009. AcMNPV as a model for baculovirus DNA replication. Virol. Sin. 24:243-267. Clarke T. E. & Clem R. J. (2003). Insect defenses against virus infection: The role of apoptosis. Int. Rev. Immunol. 22:401-424.

140

Clem, R.J. 2007. Baculoviruses and apoptosis: a diversity of genes and responses. Curr. Drug. Targets. 8:1067-1074. Cox, E.H., and McLendon, G.L. 2000. Zinc-dependent protein folding. Curr. Opin. Chem. Biol. 4:162-165. de Jong, J., 2010. Analysis of the immediate early/late protein me53 from the baculovirus Autographa californica nucleopolyhedrovirus. PhD thesis. University of Guelph, Guelph, ON. de Jong, J., Arif, B.M., Theilmann, D.A., and Krell, P.J. 2009. Autographa californica multiple nucleopolyhedrovirus me53 (ac140) is a nonessential gene required for efficient budded-virus production. J. Virol. 83:7440-7448. de Jong, J., Theilmann, D.A., Arif, B.M., and Krell, P.J. 2011. Immediate-early protein ME53 forms foci and colocalizes with GP64 and the major capsid protein VP39 at the cell membranes of Autographa californica multiple nucleopolyhedrovirus-infected cells. J. Virol. 85:9696-9707. de la Higuera, I., Ferrer-Orta, C., Moreno, E., de Ávila, A.I., Soria, M.E., Singh, K., Caridi, F., Sobrino, F., Sarafinso, S.G., Perales, C., Verdaguer, N., Domingo, E. 2018. Contribution of a multifunctional polymerase region of foot-and-mouth disease virus to lethal mutagenesis. J. Virol. pii: JVI.01119-18. doi: 10.1128/JVI.01119-18. [Epub ahead of print]. Deng F., Wang R., Fang M., Jiang Y., Xu X., Wang H., Chen X., Arif B. M., Guo L., Wang H., Hu Z. (2007). Proteomics analysis of Helicoverpa armigera single nucleocapsid nucleopolyhedrovirus identified two new occlusion-derived virus- associated proteins, HA44 and HA100. J. Virol. 81:9377-9385. Dereeper, A., Guignon, V., Blanc, G., Audic, S., Buffet, S., Buffet, S., Chevenet, F., Dufayard, J.F., Guindon, S., Lefort, V., Lescot, M., Claverie, J.M., and Gascuel, O. 2008. Phylogeny.fr: robust phylogenetic analysis for the non-specialist. Nucleic Acids Res. 36:W465-469. Dereeper, A., Audic, S., Claverie, J.M., and Blanc, G. 2010. BLAST-EXPLORER helps you building datasets for phylogenetic analysis. BMC Evol. Biol. 10:8. Du, X., and Thiem, M. 1997a. Characterization of host range factor 1 (hrf-1) expression in Lymantria dispar M nucleopolyhedrovirus- and recombinant Autographa californica M nucleopolyhedrovirus-infected IPLB-Ld652Y. Virology. 420-430. Du, X., and Thiem, M. 1997b. Responses of insect cells to baculovirus infection: protein synthesis shutdown and apoptosis. J. Virol. 71:7866-7872. Duffy, S., Shackelton, L.A., and Holmes, E.C. 2008. Rates of evolutionary change in viruses: patterns and determinants. Nat. rev. genet. 9:267-276. Faulkner, P., Kuzio, J., Williams, G.V., and Wilson, J.A. 1997. Analysis of p74, a PVD envelope protein of Autographa californica nucleopolyhedrovirus required for occlusion body infectivity in vivo. J. Gen. Virol. 78:3091-3100. Frankel, A.D., Berg, J.M., and Pabo, C.O. 1987. Metal-dependent folding of a single zinc finger from transcription factor IIIA. Proc. Natl. Acad. Sci. 84:4841-4845.

141

Gorelick, R.J., Nigida, S.M.Jr., Bess, J.W.Jr., Arthur, L.O., Henderson, L.E., and Rein, A. 1990. Noninfectious human immunodeficiency cirus type 1 mutants deficient in genomic RNA. J. Virol. 64:3207-3211. Gorelick, R.J., Chabot, D.J., Ott, D.E., Gagliardi, T.D., Rein, A., Henderson, L.E., Arthur, L.O. 1996. Genetic analysis of the zinc finger in the Moloney murine leukemia virus nucleocapsid domain: replacement of zinc-coordinating resisdues with other zinc- coordinating residues yields noninfectious particles containing genomic RNA. RNA J. Virol. 70:2593-2597. Gorelick, R.J., Gagliardi, T.D., Bosche, W.J., Wiltrout, T.A., Coren, L.V., Chabot, D.J., Lifson, J.D., Henderson, L.E, and Arthur, L.O. 1999. Strict conservation of the retroviral nucleocapsid protein zinc finger is strongly influenced by its role in viral infection processes: characterization of HIV-1 particles containing mutant nucleocapsid zinc- coordinating sequences. Virology. 256:92-104. Greenfield, N.J. 2006. Using circular dichroism spectra to estimate protein secondary structure. Nat. Protoc. 1:2876-2890. Grishin, N.V. 2001. Treble clef finger – a functionally diverse zinc-binding structural motif. Nucleic Acids Res. 29:1703-1714. Guarino, L.A., and Summers, M.D. 1987. Nucleotide sequence and temporal expression of a baculovirus regulatory gene. J. Virol. 61:2091-2099. Guarino, L.A., Xu, A.B., Jin, J., and Dong, W. 1998. A virus-encoded RNA polymerase purified from baculovirus-infected cells. J. Virol. 72:7985-7991. Guo, J., Wu, T., Anderson, J., Kane, B.F., Johnson, D.G., Gorelick, R.J., Henderson, L.E., and Levin, J.G. 2000. Zinc finger structures in the Human Immunodeficiency Virus Type 1 nucleocapsid protein facilitate efficient minus- and plus-strand transfer. J. Virol. 74:8980-8988. Hata, A., Seoane, J., Lagna, G., Montalvo, E., Hemmati-Brivanlou, A., and Massagué, J. 2000. OAZ uses distinct DNA- and protein-binding zinc fingers in separate BMP-Smad and Olf signaling pathways. Cell. 100:229-240. Herniou, E.A., Olszewski, J.A., O’Reilly, D.R., and Cory, J.S. 2004. Ancient coevolution of baculoviruses and their insect hosts. J. Virol. 78:3244-3251. Hou, D., Zhang, L., Deng, F., Fang, W., Wang, R., Liu, X., Guo, L., Rayner, S., Chen, X., Wang, H., and Hu, Z. 2012. Comparative proteomics reveal fundamental structural and functional differences between the two progeny phenotypes of a baculovirus. J. Virol. 87:829-839. Huang, J., and Levin, D.B. 2001. Expression, purification and characterization of the Spodoptera littoralis nucleopolyhedrovirus (SpliNPV) DNA polymerase and interaction with the SpliNPV non-hr origin of DNA replication. J. Gen. Virol. 82:1767-1776. Ikeda, M., Yamada, H., Hamajima, R., and Kobyashi, M. 2013. Baculovirus genes modulating intracellular innate antiviral immunity of lepidopteran insect cells. Virology. 435:1-13.

142

Imai, N., Matsuda, N., Tanaka, K., Nakano, A., Matsumoto, S., and Kang, W. 2003. Ubiquitin ligase activities of Bombyx mori nucleopolyhedrovirus RING finger proteins. J. Virol. 77:923-930. Imai, N., Matsumoto, S., and Kang, W. 2005. Formation of Bombyx mori nucleopolyhedrovirus IE2 nuclear foci is regulated by the functional domains for oligomerization and ubiquitin ligase activity. J. Gen. Virol. 86:637-644. Jehle, J.A., Blissard, G.W., Bonning, B.C., Cory, J.S., Herniou, E.A., Rohrmann, G.F., Theilmann, D.A., Thiem, S.M., and Vlak, J.M. 2006. On the classification and nomenclature of baculoviruses: a proposal for revision. Arch. Virol. 151:1257-1266. Jha, S., Rollins, M.G., Fuchs, G., Procter, D.J., Hall,E.A., Cozzolino, K., Sarnow, P., Savas, J.N., and Walsh, D. 2017. Trans-kingdom mimicry underlies ribosome customization by a poxvirus kinase. Nature. 546:651-655. Jiang, Y., Deng, F., Rayner, S., Wang, H., and Hu, Z. 2009. Evidence of a major role of GP64 in group I alphabaculovirus evolution. Virus Res. 142:85-91. Katsuma, S., Mita, K., and Shimada, T. 2007. ERK- and JNK-dependent signaling pathways contribute to Bombyx mori nucleopolyhedrovirus infection. J. Virol. 81:13700- 13709. Kaur, G., and Subramanian, S. 2016. Classification of the treble clef zinc finger: noteworthy lessons for structure and function evolution. Sci. Rep. 6:32070. Kelly, S.M., Jess, T.J., and Price, N.C. 2005. How to study proteins by circular dichroism. 1751:119-139. Klug, A., and Schwabe, J.W.R. 1995. Protein motifs 5. Zinc fingers. FASEB J. 9:597-604. Knebel-Mörsdorf, D., Kremer, A., and Jahnel, F. 1993. Baculovirus gene ME53, which contains a putative zinc-finger motif, is one of the major early-transcribed genes. J. Virol. 67:753-758. Knebel-Mörsdorf, D., Flipsen, J.T.M., Roncarati, R., Jahnel, F., Kleefsman, W.F., and Vlak, J.M. 1996. Baculovirus infection of Spodoptera exigua larvae: lacZ expression driven by promoters of early genes pe38 and me53 in larval tissue. J. Gen. Virol. 77:815- 824. Krishna, S.S., Majumdar, I., and Grishin, N.V. 2003. Structural classification of zinc fingers. Nucleic Acids Res. 31:532-550. Laitaoja, M., Valjakka, J., and Jänis, J. 2013. Zinc coordination spheres in protein structures. Inorg. Chem. 52:10982-10991. Lebrun, V., Ravanat, J., Latour, J., Sénèque, O. 2016. Near diffusion-controlled reaction of a Zn(Cys)4 zinc finger with hypochlorous acid. Chem. Sci. 7:5508-5516. Lee, G., Wu, J., Luu, P., Ghazal, P., and Flores, O. 1996. Inhibition of the association of RNA polymerase II with the preinitiation complex by a viral transcriptional repressor. Proc. Natl. Acad. Sci. USA. 93:2570-2575. Li, J., and Broyles, S.S. 1993. Recruitment of vaccinia virus RNA polymerase to an early gene promoter by the viral early transcription factor. J. Biol. Chem. 268:2773-2780.

143

Li, J.L., Cao, C., Fixsen, S.M., Young, J.M., Ono, C., Bando, H., Elde, N.C. 2015. Baculovirus protein PK2 subverts eIF2. Proc. Natl. Acad. Sci. 10.1073:E4364-E4373. Lin, G., Li, G., Granados, R.R., and Blissard, G.W. 2001. Stable cells lines expressing baculovirus P35: resistance to apoptosis and nutrient stress, and increased glycoprotein secretion. In Vitro Cell. Dev. Biol. Anim. 37:293-302. Liu, Y. 2015. Functional analysis of the baculovirus AcMNPV me53/ME53. PhD thesis. University of Guelph, Guelph, ON. Liu, Y., de Jong, J., Nagy, É., Theilmann, D.A., and Krell, P.J. 2016. Nuclear translocation sequence and region in Autographa californica multiple nucleopolyhedrovirus ME53 that are important for optimal baculovirus production. J. Virol. 90: 3953-3965. Long, L., Deng, Y., Yao, F., Guan, D., Feng, Y., Jiang, H., Li, X., Hu, P., Lu, X., Wang, H., Li, J., Gao, X., and Xie, D. 2014. Recruitment of phosphatase PP2A by RACK1 adaptor protein deactivates transcription factor IRF3 and limits type I interferon signaling. Immunity. 40:515-529. Lu, A., Krell, P.J., Vlak, J.M., and Rohrmannm, G.F. 1997. Baculovirus DNA replication. In: Miller L.K. (eds) The Baculoviruses. The Viruses. Springer, Boston, MA. Mace, P.D., Linke, K., Feltham, R., Schumacher, F.R., Smith, C.A., Vaux, D.L., Silke, J., and Day, C.L. 2008. Structures of the cIAP2 RING domain reveal conformational changes associated with Ubiquitin-conjugating enzyme (E2) recruitment. J. Biol. Chem. 283:31633-31640. Mackereth, C.D., Arrowsmith, C.H., Edwards, A.M., and McIntosh, L.P. 2000. Zinc-bundle structure of the essential RNA polymerase subunit RPB10 from Methanobacterium thermoautotrophicum. Proc. Natl. Acad. Sci. 97:6316-6321. Majzoub, K., Hafirassou, M.L., Meignin, C., Goto, A., Marzi, S., Fedorova, A., Verdier, Y., Vinh, J., Hoffmann, J.A., Martin, F., Baumert, T.F., and Schuster, C. 2014. RACK1 controls IRES-mediated translation of viruses. Cell. 159:1086-1095. McCormick, C., and Khaperskyy, D.A. 2017. Translation inhibition and stress granules in the antiviral immune response. Nat. Rev. Immunology. 17:647-660. Mészáros, B., Erdös, G., and Dosztányi. 2018. IUPred2A: context-dependent prediction of protein disorder as a function of redox state and protein binding. Nucleic Acids Res. 46:329-337. Miele, S.A.B., Garavaglia, M.J., Belaich, M.N., and Ghiringhelli, P.D. 2011. Baculovirus: molecular insights on their diversity and conservation. Int. J. Evol. Biol. Doi:10.4061. Mikhailov, V.S., anf Rohrmann, G.F. 2005. The redox state of the baculovirus single- stranded DNA-binding protein LEF-3 regulates its DNA binding, unwinding, and annealing activities. J. Biol. Chem. 280:29444-29453. Milks, M.L., Washburn, J.O., Willis, L.G., Volkman, L.E., and Theilmamnn, D.A. 2003. Deletion of pe38 attenuates AcMNPV genome replication, budded virus production, and virulence in Heliothis virescens. Virology. 310:224-234.

144

Miller, J., McLachlan, A.D., and Klug, A. 1985. Repetitive zinc-binding domains in the protein transcription factor IIIA from Xenopus oocytes. EMBO J. 4:1609-1614. Mochly-Rosen, D., Miller, K.G., Scheller, R.H., Khaner, H., Lopez, J., and Smith, B.L. 1992. p65 fragments, homologous to the C2 region of protein kinase C, bind to the intracellular receptors for protein kinase C. Biochemistry. 31:8120-8124. Mollet, S., Cougot, N., Wilczynska, A., Dautry, F., Kress, M., Bertrand, E., and Weil, D. 2008. Translationally repressed mRNA transiently cycles through stress granules during stress. Mol. Biol. Cell. 19:4469-4479. Mossman, K.L., Sherburne, R., Lavery, C., Duncan, J., and Smiley, J.R. 2000. Evidence that Herpes Simplex Virus VP16 is required for viral egress downstream of the initial envelopment event. J. Virol. 74:6287-6299. Nie, Y., Fang, M., Theilmann, D.A. 2011. Autographa californica multiple nucleopolyhedrovirus core gene ac92 (p33) is required for efficient budded virus production. 38:38-45. Nielsen, M.H., Flygaard, R.K., and Jenner, L.B. 2017. Structural analysis of ribosomal RACK1 and its role in translational control. Cell. Signal. 35:272-281. Nilsson, J., Sengupta, J., Frank, J., and Nissen, P. 2004. Regulation of eukaryotic translation by the RACK1 protein: a platform for signaling molecules on the ribosome. EMBO Rep. 5:1137-1141. Ono, C., Sato, M., Hitomi, T., Asano, S., Matsuura, Y., and Bando, H. 2015. Tightly regulated expression of Autographa californica multicapsid nucleopolyhedrovirus immediate early genes emerges from their interactions and possible collective behaviors. PLOS One. DOI:10.1371. Ono, Y., Fuji, T., Igarashi, K., Kuno, T., Tanaka, C., Kikkawa, U., and Nishizuka, Y. 1989. Phorbol ester binding to protein kinase C requires a cysteine-rich zinc-finger-like sequence. Proc. Natl. Acad. Sci. U.S.A. 86:4868-4871. Pace, N.J., and Weerapana, E. 2014. Zinc-binding cysteines: diverse functions and structural motifs. Biomolecules. 4:419-434. Passarelli, A.L., and Guarino, L.A. 2007. Baculovirus late and very late gene regulation. Curr. Drug Targets. 8:1103-1115. Peng, K., van Oers, M.M., Hu, A., van Lent, J.W.M., and Vlak, J.M. 2010. Baculovirus per os infectivity factors form a complex on the surface of occlusion-derived virus. J. Virol. 84:9497-9504. Perales-Calvo, J., Lezamiz, A., and Garcia-Manyes. 2015. The mehanochemistry of a structural zinc finger. J. Phys. Chem. Lett. 6:3335-3340. Petska, S. 1971. Inhibitors of ribosome functions. Annu. Rev. Microbiol. 25:487-562. Prikhod’ko, E.A., and Miller, L.K. 1998. Role of baculovirus IE2 and tis RING finger in cell cycle arrest. J. Virol. 72:684-692. Prikhod’ko, E.A., and Miller, L.K. 1999. The baculovirus PE38 protein augments apoptosis induced by transactivator IE1. J. Virol. 73:6691-6699.

145

Rabl, J., Leibundgut, M., Ataide, S.F., Haag, A., and Ban, N. 2011. Crystal structure of the eukaryotic 40S ribosomal subunit in complex with initiation factor 1. Science. 331:730- 736. Reed, L.J., and Muench, H. 1938. A simple method of estimating fifty per cent endpoints. Am. J. Hyg. 27:493-497. Rohrmann, G.F. 2013. Baculovirus molecular biology. Third edition. Salem, T.Z., Allam, W.R., Thiem, S.M. 2014. Verifying the stability of selected genes for normalization in qPCR experiments of Spodoptera frugiperda cells during AcMNPV infection. PLOS One. 9:e108516. Sanchez-aparicio, M.T., Rosas, M.F., and Sobrino, F. 2013. Characterization of a nuclear localization signal in the foot-and-mouth disease virus polymerase. Virology. 444:203- 210. Schultz, K.L.W., Wetter, J.A., Fiore, D.C., and Friesen, P.D. 2009. Transactivator IE1 is required for baculovirus early replication events that trigger apoptosis in permissive and nonpermissive cells. J. Virol. 83:262-272. Sekimata, M., Takahashi, A., Murakami-Sekimata, A., and Homma, Y. 2001. Involvement of a novel zinc finger protein, MIZF, in transcriptional repression by interacting with a methyl-CpG-binding protein, MBD2. J. Biol. Chem. 276:426232-42638. Sievers, F., Wilm, A., Dineen, D., Gibson, T.J., Karplus, K., Li, W., Lopez, R., McWilliam, H., Remmert, M., Söding, J., Thompson, J.D., and Higgins, D.G. 2011. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7:539. South, T.L., Blake, P.R., Sowder, R.C., Arthur, L.O., Henderson, L.E., and Summers, M.F. 1990. The nucleocapsid protein isolated from HIV-1 particles binds zinc and forms retroviral-typr zinc fingers. Biochemistry. 29:7786-7789. Tal-Singer, R., Pichyangkura, R., Chung, E., Lasner, T.M., Randazzo, B.P., Trojanowski, J.Q., Fraser, N.W., and Triezenberg, S.J. 1999. The transcriptional activaition domain of VP16 is required for efficient infection and establishment of latency by HSV-1 in the murine peripheral and central nervous systems. Virology. 259:20-33. Tao, X.Y., Choi, J.Y., Kim, W.J., Lee, J.H., Liu, Q., Kim, S.E., An, S.B., Lee, S.H., Woo, S.D., Jin, B.R., and Je, Y.H. 2013. The Autographa californica multiple nucleopolyhedrovirus ORF78 is essential for budded virus production and general occlusion body formation. J. Virol. 87:8441-8450. Taylor, S., Wakem, M., Dijkman, G., Alsarraj, M., and Nguyen, M. 2010. A practical approach to WT-qPCR – publishing data that conform to the MIQE guidelines. Methods. 50:S1-S5. Theilmann, D.A., Willis, L.G., Bosch, B.J., Forsythe, I.J., and Li, Q. 2001. The baculovirus transcriptional transactivator ie0 produces multiple products by internal initiation of translation. Virology. 290:211-223. Theilmann, D., and blissard, G., 2008. Baculoviruses: Molecular biology of nucleopolyhedroviruses. Encyclopedia of Virology. 254-265.

146

Thiem, S.M., and Chejanovsky, N. 2004. The role of baculovirus apoptotic suppressors in AcMNPV-mediated translation arrest in Ld652Y cells. Virology. 319:292-305. Thiem, S.M., Du, X., Quentin, M.E., Berner, M.M. 1996. Identification of baculovirus gene that promotes Autographa californica nuclear polyhedrosis virus replication in a nonpermissive insect cell line. J. Virol. 70:2221-2229. Tung, H., Wei, S., Lo, H., and Chao, Y. 2016. Baculovirus IE2 stimulates the expression of heat shock proteins in insect and mammalian cells to facilitate its proper functioning. PLOS One. DOI:10.1371. van Beek, N., and Davis, D.C. 2007. Baculovirus insecticide production in insect larvae. Methods Mol. Biol. 388:367-378. van Oers, M.M. 2010. Insect cell culture manual. Oxford expression technologies.1-16. van Oers, M.M., Doitsidou, M., Thomas, A.A.M. de Maagd, R.A., and Vlak, J.M. 2003. Translation of both 5’TOP and non-TOP host mRNAs continues into the late phase of Baculovirus infection. Insect Mol. Biol. 12:75-84. Vanarsdall, A.L., Mikhailov, V.A., and Rohrmann, G.F. 2007. Baculovirus DNA replication and processing. Curr. Drug Targets. 8:1096-1102. Walsh, D. 2017. Poxviruses: slipping and sliding through transcription and translation. PLoS Pathog. 16:e1006634. Wang, R., Deng, F., Hou, D., Zhao, Y., Guo, L., Wang, H., and Hu, Z. 2010. Proteomics of the Autographa californica multiple nucleopolyhedrovirus budded virions. J. Virol. 84:7233-7242. Wiedmann, C., Bellstedt, P., and Görlach. 2013. CAPITO – a web server-based analysis and plotting tool for circular dichroism data. Bioinformatics. 29:1750-1757. Wu, W., Clem, R.J., Rohrmann, G.F., and Passarelli, A.L. 2013. The baculovirus sulfhydryl oxidase Ac92 (P33) interacts with the Spodoptera frugiperda P53 protein and oxidizes it in vitro. Virology. 447:197-207. Xiao, H., and Qi, Y. 2007. Genome sequence of Lecania seperata nucleopolyhedrovirus. Virus Genes. 35:845-856. Xing, K., Deng, R., Wang, J., Feng, J., Huang, M., and Wang, X. 2005. Analysis and prediction of baculovirus promoter sequences. Virus Res. 113:64-71. Zhang, Z., Burch, P.E., Cooney, A.J., Lanz, R.B., Pereira, F.A., Wu, J., Gibbs, R.A., Weinstock, G., and Wheeler, D.A. 2004. Genomic analysis of the nuclear receptor family: new insights into structure, regulation, and evolution from the rat genome. Genome Res. 14:580-590.