Thomas Hennig

Function and transport of a herpesvirus encoded Ubiquitin-specific protease in virus entry and assembly

A thesis submitted to the degree of Doctor of Philosophy Imperial College London, Department of Medicine by

Thomas Hennig

Supervised by Professor Peter F. O’Hare Supported by Imperial College London and the Wellcome Trust September 2011 - June 2015

- 1 -

Acknowledgements

Foremost, I would like to thank Professor Peter O’Hare for his support throughout my PhD and during the post-doctoral application process. There was not a single day when you could not ask him a question or discuss results. I think I speak for everyone in the group (and outside) in that he was always approachable, patient and that I felt well supported at every point during my PhD. The last four years working with present and past members of the group have been a great experience, allowed me to learn many new techniques and made the learning process more manageable. I even started appreciating the calming influence of classical music. I would also like to thank Dr Fernando Abaitua for his help during my first year and the very detailed discussions during the early morning hours, even after having left our group over two years ago. A big ‘thank you’ and ‘sorry’ also goes to Dr Sonia Barbosa, who could usually not escape these elaborate discussions in a timely fashion, to Dr Remi Serwa, who performed the mass spectrometry, and Nora Schmidt, who had the pleasure of reading and commenting on my thesis. I was supported by two very helpful assessors, Dr Goedele Maertens and Professor Wendy Barclay, who were always happy to advise me on any issues that arose from the work presented here. Dr Maertens kindly provided the retrovirus backbones which enabled the creation of the two inducible cell lines used for the identification of VP1-2 interacting proteins. This work would have also taken much longer without the help of Dr Anna Stockum (Dr Maertens group) who helped with the retrovirus infections. During the write-up period I also received serveral helpful comments on my thesis by my PhD examiners for which I am very grateful. There is an extensive list of additional people both at St Mary’s and outside whose help should be acknowledged here but they cannot all be listed here. I am eternally grateful for all that they have done to make my work easier and overall a very pleasant experience. I would like to thank Professor Søren Paludan for enabling a short exchange to his laboratory at Aarhus University to learn valuable skills and for discussing and supporting my application to the Wellcome Trust for further funding. Finally, I was lucky enough to be funded by the Wellcome Trust, which enabled me to present and publish some of my data and attend the laboratory of Professor Søren Paludan in Aarhus (Denmark) through provision of generous grants. Substantial additional funds towards travel and accommodation were granted by the IHW and the SGM.

- 2 -

Declarations

Declaration of Originality

I confirm that this thesis has been written and the work conducted by myself unless collaborators were acknowledged. This thesis contains the work on which I expect to be examined for the degree of Doctor of Philosophy. All work that was not conducted by myself is referenced appropriately and permissions were obtained when necessary and in accordance with Imperial College policy.

The copyright of this thesis rests with the author and is made available under a creative commons attribution non-commercial no derivatives licence. Researchers are free to copy, distribute or transmit that thesis on the conditions that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work.

- 3 -

Abstract

Herpes simplex virus type I (HSV-1), the prototype α-herpesvirus (HV), is a double stranded DNA virus that replicates in the nucleus of infected cells. The nuclear pore represents a gateway that must be engaged and navigated immediately after cell entry for successful infection by many classes of human viruses. For herpesviruses, capsid-tegument assemblies must be targeted to the pore where the viral genome exits and transport into the nucleus occurs. We currently have little mechanistic knowledge of this fundamental step of infection. A swath of evidence indicates that the conserved tegument protein VP1-2 is essential for early capsid transport and pore binding, and that it contains a conserved nuclear localisation signal (NLS) required for pore docking. In this thesis I undertook a detailed analysis to dissect functional determinants within the NLS from herpes simplex virus, to examine putative NLSs in VP1-2 homologues from representatives of all sub-families, to characterise protein interactions with VP1-2 and finally to construct a GFP expressing entry defective recombinant virus to study the consequences of infection. I show that the HSV NLS can function as a mono-or bipartite motif and has a particular organisation conserved in the -herpesvirus homologues but distinct from those in the - and - herpesviruses. The representatives of all 3 classes contain a functional NLS at approximately the same position. All bi-partite motifs were able to rescue the HSV VP1-2ΔNLS virus defect albeit to varying extent whereas the mono-partite HHV-8 motif did not. I constructed and purified chimeric recombinant viruses for the VZV, HCMV and EBV motifs and show distinct differences in their ability to replicate in non-complementing cells. In HSV, NLS function in the context of protein nuclear import or of viable virus replication, was dependent on lysine 428 and the integrity of the full bi-partite motif. Mutations which reduced NLS activity generally caused reduced fitness of recombinant viruses. For the analysis of interaction partners of VP1-2 and the NLS, I developed a one-step approach to analyse the capsid interactome during entry. Additionally, I constructed mammalian GST-VP1-2.NLS fusion proteins and cell lines which inducibly express the N-terminal region of VP1-2 for analysis of NLS interacting proteins. Using mass spectrometry (MS) I identified a number of VP1-2-interacting, cellular proteins including DTX3L, an important regulator of the DNA damage response.

- 4 -

Abbreviations

BES N-bis(2-hydroxyethyl)-2-aminoethane sulfonic acid BHV-1 bovine Herpesvirus-1 BoHV-4 bovine Herpesvirus-4 bpNLS bi-partite NLS BSA bovine serum albumin CA Capsid protein (of HIV) cNLS classical NLS CPE cytopathic effect CV Crystal violet DAB Diaminobenzidine DABCO diazabicyclooctane DAPI 4′,6-Diamidin-2-phenylindol ddH2O double distilled H2O DDR DNA damage response DMEM Dulbecco's Modified Eagle's Medium DMSO dimethyl sulfoxide dsDNA/RNA double stranded DNA/RNA EBV Epstein Barr Virus EHV equine Herpesvirus EM electron microscopy ER endoplasmatic reticulum ERAD ER associated protein degradation FBS fetal bovine serum GAG glycosaminoglycan gD (or similar) HV glycoprotein D GST Glutathione-S transferase HA hemagglutinin HBS HEPES buffered saline HCMV human Cytomegalovirus HHV human Herpesvirus HIV human immunodeficiency virus HP1 heterochromatin protein 1 HR homologous recombination HRP horseradish peroxidase HS human serum HSV herpes simplex virus HV herpesvirus HVEM HV entry mediator HVS Herpesvirus Saimiri IAV influenza A virus IBB Importin-β binding domain ICP infected cell protein IN integrase (of HIV) INM inner nuclear membrane

- 5 -

Abbreviations

IP immunoprecipitation ISG interferon stimulated genes lTag large T antigen (of SV40) M1/2 matrix protein 1/2 (of IAV) MCS multiple cloning site MLV murine leukemia virus MOI multiplicity of infection mpNLS mono-partite NSL MS mass spectrometry MTOC microtubule-organising centre MuHV murid Herpesvirus MyoNLS NLS of Myopodin NCS new-born calf serum NE nuclear envelope NEAA non-essential amino acids NEBD nuclear envelope breakdown NES nuclear export signal NHEJ non-homologous end joining NLS nuclear localisation signal NPC nuclear pore complex NP nucleoprotein (of IAV) Nups Nucleoproteins ONM outer nuclear membrane ORF open reading frame PAA phosphono acetic acid PBS phosphate buffered saline Pen/Strep Penicillin/Streptomycin PFA paraformaldehyde PFU plaque forming unit PIC pre-integration complex (of HIV) PMSF Phenylmethanesulfonyl fluoride PRV Pseudorabies Virus pUL37 protein product of open reading frame 37 of unique long genome segment pUS3 protein product of open reading frame 3 of the unique short genome segment SDS sodium dodecyl sulfate SDS-PAGE SDS polyacrylamide gel electropophoresis ssDNA/RNA single stranded DNA/RNA SV40 Simian virus 40 TGN trans Golgi network ts mutant temperature-sensitive mutant USP ubiquitin-specific protease Vhs virion host shutoff protein VP virion protein Vpr viral protein R (of HIV) VZV Varicella zoster virus β-gal β-galactosidase

- 6 -

Table of contents

1 INTRODUCTION ...... - 16 -

1.1 Introduction to herpesviruses ...... - 16 - 1.1.1 Discovery and classification of Herpesviruses ...... - 16 - 1.1.2 Introduction to Herpes Simplex Virus ...... - 17 - 1.1.3 Virion structure...... - 17 - 1.1.4 The HSV-1 tegument layer ...... - 18 - 1.1.5 The HV life cycle ...... - 19 -

1.2 Properties of VP1-2 ...... - 22 - 1.2.1 Features and functions of VP1-2 ...... - 22 - 1.2.2 Role of VP1-2 during entry ...... - 23 - 1.2.3 Role of VP1-2 in replication and assembly ...... - 27 -

1.3 The nuclear import machinery ...... - 29 - 1.3.1 The nuclear membrane and nuclear pore architecture ...... - 29 - 1.3.2 NLSs and nuclear import receptors ...... - 33 - 1.3.3 RanGTP cycle and mechanism of nuclear import ...... - 36 -

1.4 Viral exploitation and modulation of nucleo-cytoplasmic transport ...... - 39 - 1.4.1 Adenovirus ...... - 44 - 1.4.2 HIV ...... - 44 - 1.4.3 Influenza ...... - 46 - 1.4.4 Baculovirus ...... - 47 - 1.4.5 Viral subversion of nucleocytoplasmic transport ...... - 47 -

1.5 Aims of PhD project ...... - 54 -

2 MATERIALS AND METHODS ...... - 58 -

2.1 Materials ...... - 58 - 2.2 Cell lines ...... - 61 - 2.3 Competent E. coli and transformation ...... - 61 - 2.4 PCR ...... - 62 - 2.5 Plasmid preparation ...... - 62 - 2.6 Transfection ...... - 62 - 2.6.1 Calcium phosphate protocol ...... - 62 - 2.6.2 GeneJammer protocol ...... - 63 -

- 7 -

Table of contents

2.7 SDS PAGE, Western blot, Coomassie and silver stain procedure ...... - 64 - 2.8 Immunofluorescence microscopy ...... - 64 - 2.9 Generation of plasmids ...... - 65 - 2.9.1 NLS β-galactosidase constructs ...... - 65 - 2.9.2 Creation of NT6 constructs ...... - 66 - 2.9.3 Creation of mammalian GST constructs ...... - 66 -

2.10 Nuclear import assay (β-galactosidase, GFP and VP1-2 constructs) ...... - 66 - 2.11 Creation of inducible NT3 and NT3ΔNLS cell lines ...... - 67 - 2.12 Immunoprecipitations ...... - 68 - 2.13 On bead digestion of IPs for mass spectrometric analysis ...... - 69 - 2.14 In silico protein analysis ...... - 69 - 2.14.1 Protein alignments, NLS, NES and Ubiquitination prediction ...... - 69 - 2.14.2 Analysis of MS data ...... - 69 -

2.15 DNA damage assay ...... - 70 - 2.16 Protocol for virus culture and infections ...... - 71 - 2.16.1 Virus infection ...... - 71 - 2.16.2 Plaque assays ...... - 71 - 2.16.3 Single step growth curve and protein expression analysis ...... - 71 - 2.16.4 Multi step growth curve ...... - 72 - 2.16.5 Virus stock production ...... - 72 - 2.16.6 Viral genomic DNA production ...... - 72 - 2.16.7 K.VP1-2.ΔNLS virus rescue assay ...... - 73 - 2.16.8 Generation of VP16-GFP and GFP-VP26 recombinant viruses ...... - 73 - 2.16.9 Time lapse microscopy of K.VP1-2.ΔNLS.VP16-GFP infected cells ...... - 74 - 2.16.10 Capsid extraction ...... - 75 - 2.16.10.1 Preparation of clean viral stocks ...... - 75 - 2.16.10.2 Capsid extraction ...... - 75 -

3 IDENTIFYING NLSS OF VP1-2 HOMOLOGUES ...... - 77 -

3.1 Background ...... - 77 - 3.1.1 Sequence alignment of the VP1-2 NLS regions from all HV sub-families ...... - 77 - 3.1.2 In silico NLS prediction in the N-terminus of VP1-2 orthologues ...... - 80 - 3.1.3 Analysis of NLS functionality ...... - 84 -

3.2 Results ...... - 85 - 3.2.1 Development of the -galactosidase nuclear translocation assay ...... - 85 -

- 8 -

Table of contents

3.2.2 Analysis of NLS function of the VP1-2 R4 regions using the -gal system ...... - 87 - 3.2.3 Analysis of NLS function of the extended VP1-2 R1 regions using the β-gal system ...... - 89 - 3.2.4 Analysis of determinants of NLS function of HSV-1 VP1-2 using the β-gal system ...... - 92 - 3.2.5 Studying NLS function in the context of HSV 1 VP1-2 ...... - 97 -

3.2.6 Expression of chimeric VP1-21-1875 ...... - 97 - 3.2.7 Analysis of localisation of mutant NT6 ...... - 98 -

3.3 Discussion ...... - 103 - 3.3.1 The choice of cargo protein can alter the outcome of the assay ...... - 103 - 3.3.2 All VP1-2 homologues contain a NLS ...... - 104 - 3.3.3 What is the nature of the NLS? ...... - 105 - 3.3.4 An improved system for screening NLSs ...... - 107 -

4 ANALYSIS OF NLS FUNCTION DURING HSV-1 REPLICATION ...... - 110 -

4.1 Studying R1 function during HSV-1 replication ...... - 110 - 4.2 Function of homologous R1s during HSV-1 replication ...... - 112 - 4.3 Functional determinants of HSV-1 R1 required for replication ...... - 114 - 4.4 Analysing pUL37 interaction with chimeric VP1-2 ...... - 114 - 4.5 Replication characteristics of recombinant viruses ...... - 116 - 4.5.1 Analysis of virus production ...... - 116 - 4.5.2 A cell-type specific defect in replication linked to the VP1-2 NLS ...... - 116 - 4.5.3 Analysis of plaque size ...... - 119 - 4.5.4 Analysis of viral protein expression ...... - 122 -

4.6 Discussion ...... - 125 - 4.6.1 A robust assay for evaluation of NLS function in supporting virus replication ...... - 125 - 4.6.2 A NLS in VP1-2 is required but not sufficient for HSV-1 VP1-2ΔNLS rescue ...... - 126 - 4.6.3 Insertion of a foreign NLS affects viral fitness ...... - 129 - 4.6.3.1 Specific defect in Vero cells ...... - 130 - 4.6.4 Future directions ...... - 132 -

5 BIOCHEMICAL ANALYSIS OF VP1-2 BINDING PARTNERS ...... - 133 -

5.1 Using the VP1-2 N-terminus to identify relevant nuclear transport factors ...... - 133 - 5.2 Results ...... - 135 - 5.2.1 Protein interaction study using the VP1-2 N-terminus ...... - 135 - 5.2.1.1 Creating stable cell lines expressing VP1-2.NT3 ...... - 135 - 5.2.1.2 Identification of VP1-2 interacting proteins: Part I ...... - 138 -

- 9 -

Table of contents

5.2.1.3 Identification of VP1-2 interacting proteins: Part II ...... - 140 - 5.2.1.4 Localisation and functions of proteins that were enriched compared to the parental control (categories I and II) ...... - 141 - 5.2.1.5 NT3 interacts with DTX3L in HEK293 cells ...... - 150 - 5.2.1.6 NT3 abolishes clustering of 53BP1 in response to the DNA damaging agent doxorubicin... - 152 - 5.2.2 Development of a system to identify cellular binding partners of incoming viral capsids- 155 - 5.2.2.1 Possible approaches to identify capsid-binding nuclear transport factors ...... - 155 - 5.2.2.2 Effect of salt concentration on capsid constitution ...... - 157 - 5.2.2.3 Effect of reducing milieu on capsid consitution ...... - 161 - 5.2.2.4 Effect of cellular proteins on capsid constitution ...... - 161 - 5.2.2.5 Effect of potassium on capsid composition ...... - 163 -

5.3 Discussion ...... - 165 - 5.3.1 Identification of VP1-2 interacting proteins using NT3 ...... - 165 - 5.3.1.1 Technical considerations for the IP ...... - 165 - 5.3.1.2 NT3 interacting proteins ...... - 167 - 5.3.1.3 NT3 interacts with DTX3L ...... - 168 - 5.3.1.4 Proposed activity of VP1-2 in altering the cellular DDR...... - 169 - 5.3.2 Identification of cellular capsid binding proteins ...... - 173 -

6 CONSTRUCTING A VP1-2 ΔNLS MUTANT EXPRESSING GFP-VP16 TO STUDY INFECTED-CELL FATE ...... - 177 -

6.1 Background ...... - 177 - 6.2 Results ...... - 177 - 6.2.1 ΔNLS.GFP-VP16 replication is modestly affected by the presence of GFP ...... - 177 - 6.2.2 ΔNLS.GFP-VP16 virus does not spread in normal cells ...... - 180 - 6.2.3 Using NLS.GFP-VP16 as a tool to study the infected cell fate ...... - 182 -

6.3 Discussion ...... - 183 - 6.3.1 Viral fitness was impaired by GFP-VP16 ...... - 183 - 6.3.2 Cluster formation ...... - 184 - 6.3.3 Proposed experiments to solve the question of cluster formation...... - 190 -

7 GENERAL DISCUSSION ...... 192

7.1 Summary ...... 192

- 10 -

Table of contents

7.1.1 Nuclear localisation and viral replication ...... - 193 - 7.1.2 Interaction studies ...... - 194 -

7.2 Discussion ...... - 195 - 7.2.1 Conclusions ...... - 195 - 7.2.2 Model of nuclear docking ...... - 196 - 7.2.3 Wider impact of NLS study ...... - 196 - 7.2.4 Some intriguing questions remain ...... - 197 - 7.2.4.1 NLS function ...... - 197 - 7.2.4.2 The expanding roles of VP1-2 ...... - 198 -

8 REFERENCES ...... - 200 - 9 APPENDIX 1: LIST OF PRIMERS ...... - 221 - 10 APPENDIX 2: ALIGNMENT OF R1 REGIONS OF HSV-1 STRAINS ...... - 227 - 11 APPENDIX 3: MS RESULTS FOR EXCISED BANDS FROM NT3 IP ...... - 228 - 12 APPENDIX 4: CATEGORY III (PROTEINS AT LEAST FOUR FOLD ENRICHED VERSUS CONTROL) - 229 - 13 APPENDIX 5: CATEGORY IV (PROTEINS AT LEAST FOUR FOLD ENRICHED) ...... - 235 - 14 APPENDIX 6: CATEGORY V (PROTEINS AT LEAST FOUR FOLD ENRICHED) ...... - 237 - 15 APPENDIX 7: ATTACHED MATERIALS ...... - 243 - 16 APPENDIX 8: PERMISSIONS FOR REPRODUCTION OF PUBLISHED MATERIALS ... - 244 -

- 11 -

Table of figures

Figure 1 Structural representation of HV features...... - 18 - Figure 2 Simplified cartoon of the HV life cycle...... - 21 - Figure 3 Simplified schematic of conserved VP1-2 domain structure of representative HVs of all sub-families. - 23 - Figure 4 Schematic representation of the NE and NPCs of metazoan cells...... - 32 - Figure 5 Structural representation of importin proteins...... - 36 - Figure 6 Retrovirus vector for creation of pTH111 (NT3) and pTH112 (NT3ΔNLS). (Maertens et al. 2010) .. - 68 - Figure 7 Flow chart to create and isolate a recombinant HSV-1 virus...... - 74 - Figure 8 Alignment of putative NLS regions of the VP1-2 N-terminus of representatives of the different HV sub-families...... - 79 - Figure 9 Development of the nuclear translocation assay; comparing GFP and β-gal NLS fusions...... - 86 - Figure 10 Localisation of β-gal fusion proteins comprising R4 homologous regions from different HV homologues...... - 88 - Figure 11 Localisation of β-gal fusion proteins featuring R1s from HV VP1-2 homologues...... - 90 - Figure 12 Analysis of a basic region in HHV-8 VP1-2 downstream of R1...... - 92 - Figure 13 Effect of point mutations in the VP1-2 NLS of HSV-1 and VZV on localisation of -gal fusion proteins...... - 95 - Figure 14 (A) Cloning strategy to create chimeric NT6 derivatives and their expression in HuH7 cells...... - 98 - Figure 15 Localisation of NT6 proteins (described in Figure 14) in HuH7 cells...... - 102 - Figure 16 Schematic representation of the ΔNLS rescue assay...... - 111 - Figure 17 Rescue assay quantification and characterisation of VP1-2 integrity...... - 113 - Figure 18 Multi- and single-step yield of recombinant HSV-1 from four different cell lines...... - 118 - Figure 19 Plaque size comparison of the recombinant HSV-1 mutants on different cell lines...... - 122 - Figure 20 Viral protein expression of ΔNLS.R (wild type) and VZV.R1 in Vero and RSC cells...... - 124 - Figure 21 Overview of the strategy to identify binding partners of the N-terminus of VP1-2 and the NLS domain...... - 134 - Figure 22 Characterisation of transduced HEK293 TetON cell clones...... - 137 - Figure 23 Small scale immunoprecipitation of NT3 and NT3ΔNLS to identify VP1-2 interacting proteins. - 139 - Figure 24 NT3 and NT3ΔNLS IPs and summary of co-precipitated proteins identified by MS...... - 140 - Figure 25 Simplified overview of annotated localisation and functions of NT3 binding proteins ...... - 142 - Figure 26 Western blot images showing the results from reciprocal co-immunoprecipitations of NT3 (A-C) and DTX3L (D-F)...... - 151 - Figure 27 53BP1 foci formation in NT3 expressing cells after doxorubicin treatment...... - 153 - Figure 28 Effects of time, salt concentration, reducing milieu and cellular proteins on tegument and capsid components...... - 160 - Figure 29 Effects of different extraction conditions on the protein composition of capsids...... - 162 -

- 12 -

Table of figures

Figure 30 Effect of high stringency extraction and purification, and potassium on virion extraction...... - 164 - Figure 31 Simplified schematic of one of the pathways leading to 53BP1 activation and how HSV-1 could potentially modulate this part of the DDR...... - 170 - Figure 32 Simplified schematic of protein levels and PARP1 activity during infection...... - 172 - Figure 33 Characterisation of ΔNLS.GFP-VP16 recombinant virus...... - 179 - Figure 34 Analysis of spread and plaque formation of ΔNLS.VP16-GFP recombiant virus over time...... - 181 -

- 13 -

Table of tables

Table 1 Overview of the consensus sequences of the six classes of classical NLSs found in eukaryotic, multicellular systems...... - 34 - Table 2 Overview of how selected viruses use and/or modify the cellular nuclear transport/pore machinery or the NE itself to enter cells...... - 41 - Table 3 Overview of how selected viruses subvert the nuclear transport machinery to facilitate efficient replication and virion egress...... - 50 - Table 4 Chemicals used for thesis. Most chemical compounds were obtained from Sigma or VWR unless otherwise stated...... - 58 - Table 5 Antibodies used for this thesis...... - 60 - Table 6 NLS prediction of two proteins that contain a confirmed classical mpNLS (SV40) or a confirmed bpNLS (nucleoplasmin)...... - 82 - Table 7 Overview of NLS predictions in the N-terminus of VP1-2 homologues using five different, freely available NLS prediction algorithms...... - 83 - Table 8 Selected Proteins that bound NT3 >four-fold better than NT3ΔNLS and parental control (Category III)...... - 145 - Table 9 Selected proteins that bound NT3ΔNLS >4-fold more than both NT3 and control (Category IV). . - 147 - Table 10 Proteins that were >4-fold enriched in both NT3 and NT3ΔNLS samples compared to parental (Category V)...... - 148 - Table 11 Localisation and presence of NLSs in documented capsid and tegument proteins...... - 187 - Table 12 Summary of the functionality of the homologous NLS regions. Efficiency is indicated compared to wild type virus (+++ same as wild type, - demarcates negative result)...... - 193 - Table 13 Summary of the functional determinants of the HSV-1 NLS region. Efficiency is indicated compared to wild type virus. (+++ same as wild type, - demarcates negative result) ...... - 194 -

- 14 -

Introduction

- 15 -

Introduction

1 Introduction 1.1 Introduction to herpesviruses 1.1.1 Discovery and classification of Herpesviruses

Herpesviruses (HVs) have plagued the human race for the whole of its existence, having evolved from ancestral species with their human hosts. Sores likely attributable to herpes simplex virus (HSV) infection have been medically described as early as 3500 years ago, even though it was only established that herpetic lesions were caused by an infectious agent in 1873. Another 50 years passed before HSV was successfully isolated and used for animal studies (Singh & Ruzek 2013). HVs are large double stranded DNA (dsDNA) viruses (Baltimore group 1) and include all members of the Herpesvirales of which more than 200 have already been described. The phylogenetic tree of HVs is expanding rapidly with the improvement of molecular genetic tools. The crudest way to classify HVs is on basis of their structure and life cycle includes all HVs, which was used to define the order Herpesvirales. They share icosahedral capsid structure and a similar mode of genome packaging. A closer look at the capsid and DNA replication (e.g. DNA polymerase) reveals similarities to Bacteriophages, which thus are assumed to have a common ancient relative (Trus et al. 2004; Casjens & King 1975; Baker et al. 2005; Pellett & Roizman 2013). Protein alignments and sequence comparisons add the detail further along the tree to allow for subdivision into three major families, which infect mammals and birds (Herpesviridae), amphibians and fish (Alloherpesviridae) and invertebrates (Malacoherpesviridae). The available sequence data and model of co-evolution with their host suggest that these might have diverged from a common ancestor between 400-1000 million years ago. Thus far over 50 genomes have been fully sequenced and over 200 different species of HVs identified (Davison 2002; Pellett & Roizman 2013). The mammalian, avian and reptilian HVs can be further divided into three sub-families, α-, β- and γ-herpesvirinae, which share between 30-43 core function genes collectively facilitating basic functions such as capsid morphogenesis, DNA encapsidation, DNA replication, entry and egress (Wang et al. 2007; Davison 2002). Many animal species were shown to be infected with at least one HV species and HV counterparts between hosts were identified. Thus it is no surprise that the list of HVs is predicted to expand through the use of rapidly evolving DNA amplification and sequencing methods (Pellett & Roizman 2013). The sub-families are believed to have branched around 200 million years ago and most of the currently sequenced HV genomes are classified as belonging to this branch. Besides inhabiting similar niches within their hosts, sub-families share some genetic features, that is, chunks in the central region of their genomes, which carry core genes, show similar gene arrangement and content. The nine human HVs (HHV-1 through 8, with HHV-6 divided into A/B types) span all three sub-

- 16 -

Introduction families and their life cycle shows sub-family specific adaptations. While α-HVs can infect and establish latency in neurons, the β- and γ-HVs can form reservoirs in myeloid-derived cells (amongst others) (Pellett & Roizman 2013).

1.1.2 Introduction to Herpes Simplex Virus

HHVs have been associated with trivial ailments such as orolabial lesions, chicken pox and mononucleosis, and more severe sequelae such as encephalitis, Guillain Barre Syndrome, keratitis and post-transplant disseminated infection but quite often they cause no discernible symptoms at all (Roizman et al. 2013). HSV-1 is the prototypical and best described of the human α-HVs and lends itself to many morphological and replication studies since it is, at this point in time, easier to grow than most of its relatives. HSV-1 is a promiscuous virus infecting many different cell types in vitro but in vivo infection usually occurs in epithelial cells leading to characteristic pathology including orolabial lesions or keratitis. Generally, disease is recurrent and self-limiting but can, in immunocompromised and neonatal individuals, lead to disseminated infection and encephalitis (Pellett & Roizman 2013; Roizman et al. 2013). Epidemiological estimates from the USA indicated that HSV-1 infects up to 60% of the population (Xu et al. 2002; Xu et al. 2006), which may vary depending on ethnicity and socioeconomic status and that it is transmitted through close contact with saliva or exudates and initiates infection usually in mucosal sites (Roizman 2011).

1.1.3 Virion structure

All HVs share common structural features (Figure 1A), each featuring a host cell derived lipid envelope which is attached to the viral nucleocapsid via an amorphous protein network termed the tegument. The linear dsDNA genome of HHVs ranges in size from 125-236kb and is enclosed under high pressure to enable expulsion upon opening of the portal. The capsids of all HVs are icosahedrons (T=16) around 100-125nm across which are formed by 150 hexameric and eleven pentameric capsomeres which are made up of 955 copies of the 170kDa virion protein (VP) 5 and one portal (12 copies of portal protein) which is illustrated in Figure 1B. Other constituents of nucleocapsids are the 320 triplexes, made up of one copy of VP19C and two copies of VP23, which link the capsomeres (Newcomb et al. 1993), and the non-essential small capsid protein VP26. Upon assembly of the nucleocapsid the tegument layer is sequentially added to it, which enables the recruitment of an envelope during the egress phase of assembly (Coller et al. 2007).

- 17 -

Introduction

Figure 1 Structural representation of HV features. (A) Electron micrograph showing a negatively stained HV virion comprising a host derived membrane decorated with viral glycoproteins. The underlying tegument layer connects the membrane to the icosahedral capsid which encloses the double stranded viral DNA genome. (B) Reconstruction of one basic icosahedron that makes up the HV capsid (T=16). It is made up of 12 pentons and 150 hexons which are held together by triplexes. (Brown & Newcomb 2011)

1.1.4 The HSV-1 tegument layer

Tegument constituents are only partially conserved amongst the HVs and composition is determined by the virus (i.e. it is mostly cell type independent) but may vary depending on when the virus is assembled (early vs. late) (Newcomb & Brown 2009; Loret et al. 2008). Over the years the composition of the HSV-1 tegument has been serially refined (see Loret et al. 2008; Kelly et al. 2009) including references for a list of tegument components) and, depending on which technique was used, up to 24 different protein species have been described within mature virions (Loret et al. 2008; Michael et al. 2006; Newcomb et al. 2012; Henaff et al. 2013; Bohannon et al. 2013). Maybe not surprisingly, even cellular proteins (Loret et al. 2008) and RNAs (Sciortino et al. 2001) can be found in mature virions. Given that >50% of the virion mass is contributed by the tegument layer (Loret et al. 2008) it may not be unexpected that these proteins fulfil important, and sometimes essential, functions during both virus entry and establishment of the early infection programme (reviewed in Kelly et al. 2009). It is not uncommon for tegument protein to function in bi-phasic manner (both pre-early and late) although in vitro quite a few can be deleted without any discernible effect on infection or assembly (Kelly et al. 2009). Since these proteins are delivered as part of the inoculum most of them likely function prior to onset of viral transcription or to facilitate initiation of transcription either directly or indirectly. Characteristically, they also fulfil their functions in cells treated with protein synthesis inhibitors such as cycloheximide. Their confirmed functional contributions in the very early stage of infection include establishment of immediate early transcription (Infected cell protein (ICP) 4, VP16), modulation of innate cellular defences (ICP0), degradation of cellular mRNA (by Vhs (virion host shutoff protein)), and sequentially, capsid movement along microtubules, delivery to the nuclear pore and

- 18 -

Introduction uncoating of the genome (VP1-2 and protein product of open reading frame 37 of the unique long HSV- 1 genome segment (pUL37)) (Kelly et al. 2009), and later on inhibition of apoptosis (pUL14 and protein product of open reading frame 3 of the unique short genome segment (pUS3)) (Yamauchi et al. 2003; Munger & Roizman 2001), modulation of DNA damage repair (ICP0) (Wilkinson & Weller 2006), replenishment of cellular nucleotide pools (pUL23) (Chen et al. 1979), cytoskeletal rearrangement (pUL21) (Takakuwa et al. 2001), nuclear egress (pUS3) (Purves et al. 1991) and secondary envelopment (VP1-2, pUL37, VP16 etc.) (Desai 2000; Klupp et al. 2001; von Einem et al. 2006).

1.1.5 The HV life cycle

HSV-1 is a promiscuous virus replicating in a variety of cells providing it finds a suitable receptor on the cellular surface. For the purpose of this thesis the following paragraphs will highlight the major steps in HV entry, transport, nuclear entry, replication and egress during lytic infection. For additional information on latency which is beyond the scope of this thesis I refer the reader to some elegant reviews by Roizman & Whitley (2013); Nicoll et al. (2012). The initial non-essential (but nonetheless quite beneficial) steps in HV infection involves the loose binding of enveloped virions to ubiquitous, positively charged cell surface proteoglycans (glycosaminoglycans (GAGs)) via its gC glycoprotein (Herold et al. 1991). This brings the virion into closer proximity with the cell surface where it can scan, via its gD glycoprotein, for high affinity receptors including Nectin-1, HVEM (HV entry mediator) and specifically modified heparan sulphates (Figure 2, step 1) (Krummenacher et al. 1998; Nicola et al. 1998; Campadelli-Fiume et al. 2000). This interaction is, however, not sufficient for infection but allows the dimer gH/L and gB to initiate fusion events (Figure 2, step 2) (Subramanian & Geraghty 2007). While fusion can take place at the cytoplasmic membrane as noted in the earliest HSV studies over 40 years ago (Morgan et al. 1968; Fuller & Spear 1987), virions can also productively or unproductively be taken up by cells (Nicola et al. 2005; Epstein et al. 1963; Nicola & Straus 2004). The decision for immediate fusion or endocytosis depends on the cell type and thus presence of co-receptors such as specific integrins and viral determinants (Hadigal & Shukla 2013). In the absence of gD, for instance, HSV-1 was still seen to enter cells via endocytosis but infection never ensued (Nicola & Straus 2004). In both cases, the immediate post-fusion events lead to a loss of tegument structures (Figure 2, step 3) from the capsid. These are partly retained at the inner surface of the cytosolic membrane as seen by immuno-electron microscopy (EM) and eventually lost during transit to the nucleus (Granzow et al. 2005; Aggarwal et al. 2012). Both cellular and viral factors have been implicated in mediating this partial uncoating or de-tegumentation which facilitates the binding of capsids to the cellular microtubule network. These partially uncoated capsids, retaining at least VP1-2 (pUL36), pUL37 and

- 19 -

Introduction pUS3, migrate in retrograde fashion (Figure 2, step 4) towards the nucleus (Sodeik et al. 1997; Radtke et al. 2010; Wolfstein et al. 2006) via the microtubule organising centre (Abaitua et al. 2012). Once they reach a juxtanuclear position capsids dock at the nuclear pore in a process likely facilitated by one of the inner tegument components. Binding at the nuclear pore initiates a set of poorly defined events leading to DNA injection into the nucleus (Figure 2, step 5). Through a complex set of events involving tegument proteins, including ICP0, ICP4 and VP16, transcription of immediate early (α) genes is facilitated (see Figure 2, step 6) and viral genomic silencing is inhibited (Roizman 2011; Post et al. 1981). The α genes include the transcription factor ICP4 and the E3 ligase ICP0. These factors also lead to sequential expression of β and later on γ genes, the latter of which occurs only once DNA synthesis had been initiated (reviewed in Roizman et al. 2013). At this stage the pre-replicative foci forming around the incoming genomes evolve into mature replication compartments where the complex DNA replication yields concatemeric DNA intermediates which are subsequently packaged (step 7/8) (Liptak et al. 1996; Zhong & Hayward 1997; Schumacher et al. 2012; Lo Piano et al. 2011). The β and γ genes comprise all viral factors required for appropriate DNA replication and nuclear assembly of nucleocapsids at late stages (steps 7-9). The nascent capsids progress from A capsids, containing scaffolding proteins but lacking DNA, to B capsids which contain no scaffold any more but still lack DNA, to C capsids which contain besides non-nucleosomal DNA the cellular protein spermine (Tandon et al. 2015). These types of capsids were named in the order they appear if capsids obtained from infected-cell nuclei are purified using a density gradient. Studies in cell free systems and baculovirus transduced cells elegantly revealed that capsids can self-assemble in the absence of any of the α and β gene products, which are required for their expression. In this model only the small capsid protein VP26, which is also not required for virion formation in cell culture, was dispensable (Thomsen et al. 1994; Newcomb et al. 1994). Once capsids are fully packaged they embark on the egress pathway. While C capsids have been seen to traverse nuclear pores or leave the nucleus through induced membrane breakdown, overwhelming evidence lead to the formulation of the generally accepted model of primary envelopment at the inner nuclear membrane through the recruitment of several kinases by pUL31/34, which in complex can cause vesicle release into the perinuclear space even in the absence of infection (steps 9-10) (Mettenleiter et al. 2013; Schulz et al. 2015). These perinuclear virions subsequently lose their envelope through fusion at the outer nuclear membrane which is facilitated by virion associated pUS3. Unlike pUS3, the complex pUL31/34 is retained at the nuclear membrane (Henaff et al. 2013). The now cytoplasmic capsids are sequentially tegumented by inner tegument proteins, including VP1-2 and pUL37 (Henaff et al. 2013), and later on outer tegument proteins either at trans-Golgi

- 20 -

Introduction derived vesicles or endosomes via homo- and heterotypic interactions of outer tegument-outer tegument and outer tegument-glycoprotein (steps 11-13) (Mettenleiter 2006; Granzow et al. 2001; Vittone et al. 2005). The vesicularised mature virions are then transported to the plasma membrane where they are released in a process similar to exocytosis (step 14) (Harley et al. 2001; Hollinshead et al. 2012; Avitabile et al. 1995). The mature virions can now infect subsequent cells and close the cycle. Depending on the cell type and which HV was used to infect a target cell, the cell will produce, besides infectious virions, other capsid-less vesicles filled with outer tegument and cellular proteins (light particles) (Szilágyi & Cunningham 1991; Dargan et al. 1995).

Figure 2 Simplified cartoon of the HV life cycle. Details are explained in the text. Picture taken from (Https://www.fli.bund.de/typo3temp/pics/201d5a0ec0.jpg n.d.). T – tegument, M – mitochondrium, MT – microtubules, RER – rough endoplasmatic reticulum, G – Golgi apparatus, NP - nuclear pore, NM – nuclear membrane, N – nucleus, TGN – trans-Golgi network

- 21 -

Introduction

1.2 Properties of VP1-2 1.2.1 Features and functions of VP1-2

VP1-2 is the largest HV protein consisting of 3164 residues which are encoded in the UL36 open reading frame (ORF) of the HSV-1 genome (McGeoch et al. 1988). Genomic studies on other HVs indicate that it is conserved throughout all sub-families (Figure 3) but overall sequence identity is only around 30-35% (Izumiya et al. 1999). Structurally the protein can be divided into multiple universally conserved areas which are linked by sub-family specific linker regions. The N-terminus features a conserved and independently functional ubiquitin-specific protease (USP) domain (Kattenhorn et al. 2005; Schlieker et al. 2005). Interestingly, while its function is dispensable in cell culture systems, deletion of the domain prevents formation of virions suggesting a structural role during assembly. The overall sequence conservation is moderate but all homologues feature a conserved cysteine characteristic of a cysteine protease. The region downstream of the USP is highly variable in both primary sequence and length but contains a short stretch of high conservation. Secondary structure algorithms predict this region to be disordered which likely is caused by the relative enrichment of proline residues which are known to break secondary structures. In HSV, PRV (Pseudorabies virus) and HCMV (human Cytomegalovirus) this region was shown to exhibit strong nuclear localisation signal (NLS) activity (Möhl et al. 2009; Abaitua & O’Hare 2008). Results obtained by our group (Abaitua et al. 2012) and a group working on HCMV (Brock et al. 2013) showed that this NLS was essential to virus replication. Flanking the NLS lie the N- terminal VP16 (Svobodova et al. 2012) and the C-terminal UL37 binding region (Klupp et al. 2002; Mijatov et al. 2007), both of which are also essential to viral replication independently of the NLS. Directly adjacent to the UL37 binding region follows a long segment which is relatively well conserved across all HVs. Until recently not much was known about its function or structure. In silico analysis indicated the presence of several coiled-coil domains but functional significance was lacking. It was only recently that structural studies of the middle segment gave an indication as to how VP1-2 functions as part of the virion (discussed later). That study found that VP1-2 can exist in monomeric or dimeric forms forming extended fibres through a series of loop-connected helices between residues 520-1920 (Scrima et al. 2015). This segment connects to the highly conserved extreme C-terminal domain via a poorly conserved linker that seems almost entirely dispensable in cell culture during infections with another αHV, PRV (Böttcher et al. 2006). That linker features a UL25 binding domain and a PQ repeat. Additionally, in Marek’s disease virus the proline rich section is only present in the oncogenic type 1 but not the non-oncogenic type 2 (Izumiya et al. 1999). In contrast the conserved C- terminus is absolutely required for virus replication and assembly and comprises a second UL25

- 22 -

Introduction binding domain likely enabling recruitment of VP1-2 to nascent capsids (Lee et al. 2006; Coller et al. 2007).

Figure 3 Simplified schematic of conserved VP1-2 domain structure of representative HVs of all sub-families. The approximate positions of known binding sites of viral proteins in the HSV-1 VP1-2 homologue are demarcated with black arrows above. Non-conserved linker regions contain no shading and are demarcated with black arrows below the HSV-1 homologue. Shaded areas show regions (or domains) that are conserved across all sub-families. Differences in shading indicate relative conservation compared to HSV-1. Four separate conserved regions can be found in each VP1-2 homologue; the N-terminal USP (ubiquitin specific protease domain, in brown); an essential, highly basic stretch of amino acids which possesses NLS activity in at least HSV-1 (red); a relatively well conserved middle segment that likely forms a fibril (grey); the very well conserved and essential C-terminus (blue) that facilitates the VP1-2 capsid interaction of at least HSV-1. Adapted from (Hennig et al. 2014) with additional information from (Scrima et al. 2015; Fuchs et al. 2004; Lee et al. 2006; Klupp et al. 2002; Mijatov et al. 2007; Vittone et al. 2005; Pasdeloup et al. 2009; Coller et al. 2007).

1.2.2 Role of VP1-2 during entry

Evidence from over three decades ago suggested that VP1-2 is transcribed and translated after onset of DNA synthesis and that it is incorporated into mature virions isolated from cells either in the nucleus or after primary de-envelopment in the cytoplasm. As such it was consistently found as a group of up to three high molecular weight bands present in highly purified extracellular virions. From its migration properties and band intensity it was estimated to be present in between 110-150 copies and due to its resistance to extraction is likely associated with incoming capsids. In fact, it was even possible to co-precipitate capsid proteins (e.g. VP5) and, more recently, to see virion-associated VP1-2 in immuno-EM studies using a specific antiserum raised against purified VP1-2 (Gibson & Roizman 1972; Spear & Roizman 1972; Honess & Roizman 1973; Jones & Roizman 1979; Heine et al. 1974; D S McNabb & Courtney 1992; Granzow et al. 2005). The first indication that VP1-2 is essential and might be contributing to HSV entry came from experiments using a temperature-sensitive (ts) mutant, tsB7,

- 23 -

Introduction which was blocked prior to expression of immediate early genes (Figure 2, steps 3-5) but could complement the growth of other ts mutants in trans while its phenotype was largely (cis-) dominant. Although it was not known which protein(s) was affected by the mutation(s) it was sequentially mapped to a region between 0.501-0.503 map units along the HSV-1 genome using a HSV-2 marker rescue system. The authors then suspected that the tsB7 lesion likely affected the uncoating process (Figure 2, step 5) as viral titres of tsB7 did not rise congruently with a second trans-complemented mutant. The uncoating defect was later confirmed elegantly using EM which showed that DNA-filled HSV capsids accumulate in a juxtanuclear position and that DNA release was extremely slow at the non-permissive temperature (Batterson & Roizman 1983; Knipe et al. 1981; Schaffer et al. 1978). Interestingly, full capsids could still be seen at the nuclear envelope (NE) at 6hpi and still productively release their cargo upon a short shift to the permissive temperature implying a relative longevity and resistance to cellular degradation of mature capsids including any capsid-associated factor required for genome uncoating and nuclear pore attachment (i.e. that went through appropriate tegumentation and egress). This very much contrasts the observed breakdown of capsids and loss of DNA in gD-null mutant viruses that do not recruit their secondary envelope (Figure 2, steps 12/13) and highlights the distinct nature of both cytoplasmic capsid species (Campadelli-fiume et al. 1991). For a virus particle to reach the nucleus it has to cross the cytosol. Because of the high level of viscosity of the cytosol, without appropriate guidance, diffusion of capsids would take quite a long time depending on the distance of the nucleus from the cytoplasmic membrane and would be deemed unlikely in neurons. Since HSV-1 infection in vitro in non-neuronal cells was detected as early as 45mpi (Fenwick & Clark 1982) it can be assumed that migration is assisted by the cytoskeleton. Experiments from nearly three decades ago showed that this process was sensitive to microtubule depolymerising and stabilising drugs in neurons (Kristensson et al. 1986) although in Vero cells, which feature a comparatively short distance between cytoplasmic and nuclear membranes, that inhibition was less pronounced and mostly absent with the stabilising agent Taxol (Sodeik et al. 1997). Later results expanded these data through detailed microscopic investigations in PRV and HCMV infected cells which was equally dependent on microtubules and, more impressively, found capsids localising near microtubules (Granzow et al. 2005; Ogawa-Goto et al. 2003). Theoretically, any exposed capsid or tegument protein (some examples of microtubule-interacting HSV proteins reviewed in (Döhner et al. 2005)) can mediate this function and one can draw logical conclusions from several lines of telling evidence that narrow down the search for the enigmatic driver of both movement and nuclear pore docking during entry. (i) VP26, an exposed capsid component, remains attached to capsids during transit and can mediate binding to microtubules. Artificially assembled capsids that lack VP26 (VP26-null) bind less

- 24 -

Introduction well to nuclei (Douglas et al. 2004) but migration did occur without it (Döhner et al. 2006). Earlier it was already found that VP26 was non-essential in vitro and in mice although virus recovery was reduced (Desai et al. 1998). VP26-null mutants were also able to infect neurons both in vitro (Antinone et al. 2006) and, more importantly, in a mouse keratitis model (Desai et al. 1998). Since VP5 and the triplexes likely are buried under VP26 and remaining tegument structures the spotlight logically falls onto the latter. (ii) Incoming capsids only react with specific antisera to VP1-2, pUL37 and pUS3 in immuno-EM studies or can be found to retain GFP-VP1-2 or GFP-UL37 during PRV transit in neurons (Granzow et al. 2005; Luxton et al. 2005) (iii) Capsids derived from nuclei, which are devoid of tegument structures, cannot migrate along microtubules in vitro. Similarly, capsids extracted with Triton X-100 at physiological salt showed negligible migration since they are covered in both outer and inner tegument proteins, which theoretically obstructs binding to motor proteins either sterically or through the activity of a tegument protein. It was only when virions were extracted in the presence of 0.5-1M KCl, yielding capsids devoid of some outer tegument proteins but retaining most of the inner tegument, that these manage to bind and mediate capsid migration along microtubules in vitro. This phenomenon was sensitive to the cellular microtubule transport inhibitor dynamitin. It has to be noted, however, that in the study by Radtke et al. capsids treated with the higher salt concentration lost the ability to precipitate microtubule motors (Wolfstein et al. 2006; Radtke et al. 2010). Additionally these capsids looked different to those found by immuno-EM in cells in a previous study by Granzow et al. (VP16, VP22 and VP13/14 present) but they were able to migrate. One would argue that immunoblotting and MS are more sensitive and that immuno-EM relies heavily on the availability of epitopes which may be altered by any of the fixation or incubation steps involved in specimen preparation. Nonetheless these results essentially preclude capsid protein involvement in microtubule recruitment. Several factors were found to bind microtubules in vitro (in isolation) and could contribute to recruitment. The outer tegument protein pUL11 (Lyman & Enquist 2009), for example, binds microtubule components and is almost completely removed by 1M KCl which also coincides with loss of microtubule recruitment (Radtke et al. 2010). However, pUL11 is non-essential to HSV-1 and PRV replication in cell culture (Leege, Fuchs, et al. 2009). This leaves the proteins VP1-2, pUL37 and pUS3. (iv) The role for pUL37 is more complex and controversial. In PRV, for instance, UL37-null virus or an independent mutant in which VP1-2 fails to recruit pUL37 remains viable although virus titres drop substantially (Klupp et al. 2001; Fuchs et al. 2004). One study on HSV-1 identified dystonin as factor required for secondary envelopment, which was abrogated in UL37-null virus. In contrast to PRV, pUL37 is essential in HSV-1 and deletion is difficult without impairing virion formation completely

- 25 -

Introduction

(Pasdeloup et al. 2010). However, this can be elegantly circumvented using a heterokaryon system, in which uninfected cells in a monolayer of cells infected at low multiplicity of infection (MOI) are fused with infected ones. In this system wild-type and UL37-null HSV-1 were both able to infect uninfected nuclei within the boundaries of the same cytoplasmic membrane (Roberts et al. 2009). (v) In contrast a VP1-2 deficient virus was not able to spread to other nuclei within the same heterokaryon (Roberts et al. 2009). Although there appears to be conflicting evidence on the involvement of UL37 during entry this study suggests VP1-2 is the dominant factor for infectivity. (vi) A mutant with a similar phenotype, which lacks the extreme C-terminal UL25 binding site in VP1-2 was also not able to infect subsequent cells since the VP1-2-capsid association was too loose using only one of the UL25 binding domains. Incoming capsids thus lose VP1-2 during de- tegumentation (Figure 2, steps 2/3) or transit (steps 3-5) precluding successful nuclear docking (Schipke et al. 2012). (vii) A mutant virus harbouring a deletion of a highly conserved NLS within VP1-2, which did not affect virion formation or fusion with the membrane of subsequent cells, was similarly deficient for infection. This defect was traced to a defect in either routing capsids from the microtubule organising centre (MTOC) to the nucleus or docking at the nuclear pore (Abaitua & O’Hare 2008; Abaitua et al. 2012; Möhl et al. 2009). A similar defect was later confirmed in HCMV and it was shown that the HSV- 1 motif could partially rescue if the HCMV VP1-2 containing the HSV-1 NLS was expressed in trans (Brock et al. 2013). (viii) As previously described the tsB7 mutant can bind to the nuclear pore even if entry was performed at the non-permissive temperature. This phenotype was exploited in a study by Jovasevic et al. in which tsB7 capsids were left to dock at the nuclear pore complex (NPC) at the non-permissive temperature. Prior to the shift to the permissive temperature, which would allow genome uncoating and immediate early gene expression to ensue, a serine-cysteine protease inhibitor was added to the medium which in turn abolished genome release even though the temperature was lowered. The authors speculate that cleavage has to occur at the nuclear pore to facilitate a conformational switch leading to the release of the HSV-1 genome into the nucleus, since they observed a 55kDa N-terminal cleavage product only in conditions that enabled DNA injection (Jovasevic et al. 2008). (ix) A poorly conserved area of the αHV VP1-2 C-terminus. Although the VP1-2 C-terminus is poorly conserved across the sub-families a large region of 709 residues can be deleted from the PRV homologue with minimal effect on replication in cell culture (Böttcher et al. 2006), further work has shown that a proline-glutamine repeat embedded within that deletion attenuated infection of neurons in a mouse eye infection model (Zaichick et al. 2013). In order to infect neurons, capsids have to move along axons in retrograde fashion. Movement along microtubules, however, is not a simple one-way

- 26 -

Introduction process and capsids were shown to make simultaneous contact with motors of the bi-directional machinery (Radtke et al. 2010) but had preference for motion in the ‘right’ direction depending on the context. Interestingly, deletion of the PQ rich region (part of the 709 amino acid deletion) caused capsids to pause more often, effectively shortening the length of a run (guided motion in one direction). Additionally, capsids were seen to move in the ‘wrong’ direction (anterograde) more often (Zaichick et al. 2013). (x) Further downstream of the PQ motif lies a region which is also universally conserved in all HVs although there are some sequence differences between sub-families. The presented evidence essentially narrows down the list of suspects to VP1-2. While not all domains of this long protein are required for infectivity (e.g. USP) it still remains the only truly essential protein in this process. Very recent structural evidence (Scrima et al. 2015) will likely lead to more complete understanding of the effects of the above described mutations and may even give an indication how VP1-2 fulfils functions in both assembly, entry and nuclear pore docking.

1.2.3 Role of VP1-2 in replication and assembly

Along with its previous role in entry VP1-2 is indispensable during assembly of HV virions. It is thought to provide the platform for the addition of further tegument structures to nascent capsids by directly interacting with essential tegument proteins pUL37 and VP16, which both bind to the N- terminal half of the protein and likely recruit further proteins (e.g. VP16 was shown to interact with multiple tegument proteins), guide capsids to areas of secondary envelopment and prevent auto- aggregation of capsids in the cytosol (Bucks et al. 2007; Vittone et al. 2005; Mettenleiter 2006; Desai 2000). Evidence of its essential nature during assembly came from multiple studies using ts mutants, ΔUL36 mutant viruses or partial deletions. The previously described tsB7 has become a very useful tool for investigation of uncoating. Our group has previously shown that map region 0.501-0.503 corresponds to VP1-2 and is associated with exactly four substitutions. Of these the Y1453H was sufficient to recapitulate the tsB7 defect (Abaitua et al. 2009; Abaitua et al. 2011). But the mutation(s) that were introduced through crude chemical mutagens brought with it/them another defect late during replication. When infections were performed at the permissive temperature or were shifted to the permissive temperature after nuclear docking then DNA uncoating was facilitated. Immediate early and early gene expression appeared to occur normally but if the temperature is shifted back to the non-permissive temperature before DNA replication started then late gene expression was suppressed. If the temperature was shifted after the onset of late protein synthesis then, although levels of VP5 are similar to wild-type virus at late times, VP1-2 seemed to aggregate in the cytoplasm which lead to impaired progeny virus production. This points to a role for VP1-2 in assembly. The reason for the lack of late gene expression at the non-

- 27 -

Introduction permissive temperature still remains to be elucidated but may not lie within VP1-2 since tsB7 VP1-2 revertant virus still showed some defect (Abaitua et al. 2011). Another point to address would be whether VP1-2 might contribute to nuclear egress which also involves the controversial question on where it may be added to nascent capsids. VP1-2 homologues in at least PRV and HCMV contain a NLS (Brock et al. 2013; Möhl et al. 2009) and early on it was clear that VP1-2 can enter nuclei despite its large size (D. S. McNabb & Courtney 1992). Several studies found VP1-2 on nuclear capsids while others showed its absence (Radtke et al. 2010; Henaff et al. 2013). The first study in a series of VP1-2 deletion mutants where the C-terminal 2700 residues were deleted indicated that deletion of most of VP1-2 had no effect on HSV DNA encapsidation into nascent capsids and little (Desai 2000) to moderate (Luxton et al. 2006) effects of nuclear egress of HSV-1 and PRV, respectively, from Vero cell nuclei as judged from fluorescence microscopy and EM. However, there were some differences in the mutants used in both studies. The HSV-1 ΔUL36 virus expressed the USP domain and the PRV ΔUL36 mutation caused a concomitant reduction of UL37 expression. The former might alleviate the ΔUL36 egress defect by contributing USP function while in the latter scenario an essential function might be impaired as was seen in a HSV-1 ΔUL37 mutant (Desai et al. 2001). It has to be noted that Desai et al. mentioned the presence of immature capsids in the cytoplasm which contained no DNA but scaffold proteins. However, this might occur artefactually as in these cells a vast amount of capsids accumulate in nuclei in contrast to wild-type viruses. Furthermore, these capsids never acquired further tegument structures or an envelope and were diffusely localised within the cytoplasm (Desai 2000). To answer the question of whether VP1-2 is added in the nucleus could also be addressed by indirect evidence from ‘milder’ mutations. Our group has, as described above, shown that nuclear localisation of VP1-2 is entirely dependent on the NLS found in its N-terminus (Abaitua & O’Hare 2008). In this setting VP1-2 cannot enter the nucleus and, while entry of capsids was impaired concomitantly, assembly occurred essentially normally. These results might further preclude any additional essential role of VP1-2 in DNA synthesis, packaging and cleavage as hinted by ΔVP1-2 mutant since late protein synthesis and DNA encapsidation occurred quite normally (Desai 2000). Later studies on a true VP1-2-null PRV confirmed the defects in secondary envelopment (Fuchs et al. 2004) which is likely amplified in HSV-1. The gold standard to confirm where VP1-2 is added to capsids and whether it may thus play a role in capsid egress would involve the investigation of nuclear capsids. There are in fact several studies giving direct evidence that VP1-2 might not be added to capsids within the nucleus. Most evidence came from the study of nuclear capsids which are often used as negative controls for microtubule binding assays. Several groups analysed nuclear C capsids, which already contain DNA, for their VP1-2 and pUL37 status. By EM characteristic tufts extending from pentons are observed in in vitro de-

- 28 -

Introduction tegumented capsids but not C capsids. This was confirmed by immunoblots against VP1-2 in the SDS lysates of purified nuclear C capsids derived from infected cells (Newcomb & Brown 2010; Radtke et al. 2010; Wolfstein et al. 2006). While this evidence appears conclusive there are studies that contradict these findings (Lee et al. 2006; Henaff et al. 2013; Bucks et al. 2007; Leelawong et al. 2012). The discrepancy in the data might have resulted from the different systems used to study VP1-2 recruitment. Some studies use high MOI infections of Hela S3 cells with early lysis while other use low MOI multi round infections of BHK-1 cells which might show slightly altered assembly steps and localisation patterns of VP1-2. Furthermore at very late time points the cells will have suffered a lot from the ongoing cell death, mediator release and general disintegration of the infected cells that nuclear transport may have been impaired. Conversely, high MOI infection may also affect nuclear transport of VP1-2 or alter the spatiotemporal pattern of assembly and early tegumentation. While the role for nuclear VP1-2 is controversial, evidence for later stages appears to unanimously implicate VP1-2 in directing capsids to sites of secondary envelopment (Desai 2000; Desai et al. 2001; Klupp et al. 2001). Additionally, VP1-2 is both sufficient and essential for the subsequent recruitment of other inner and outer tegument structures both directly and indirectly (Bucks et al. 2011; Mijatov et al. 2007; Klupp et al. 2002). How is it possible for VP1-2 to link capsids to the envelope? Evidence suggests that VP1-2 extends outwards from pentons in characteristic ‘tufts’ of triton X-100 extracted capsids. These structures appear 30-50nm in length as estimated from EM studies (Newcomb & Brown 2010) which would likely be sufficient to span the tegument. This way it could serve as an assembly platform for, initially, pUL37 and VP16 which both make contacts with subsequent proteins (Vittone et al. 2005). Recently, a study found that the VP1-2 central region between residues 520-1920 could form a stalk assembly and may facilitate homo-dimerisation of VP1-2 into a superhelix which is sufficiently long to span the tegument layer (Scrima et al. 2015). From this evidence combined with EM capsid studies and the fact that deletion of VP1-2 abolishes all tegumentation leads to the momentary conclusion that VP1-2 serves as a bridge between capsid and envelope as suggested previously.

1.3 The nuclear import machinery 1.3.1 The nuclear membrane and nuclear pore architecture

The NE (NE) is a physical barrier that separates the cytoplasmic contents from the genetic material. This spatial separation affords cells increased possibilities to regulate their life cycle and differentiation state with high complexity which forms the basis of multicellular organisms. This barrier is made up of a double membrane and a tight interstitial space, the perinuclear space. The outer

- 29 -

Introduction nuclear membrane (ONM) is contiguous with the endoplasmatic reticulum (ER) which in theory allows proteins from the perinuclear space to access the secretory pathway and vice versa. There are several thousand nuclear pores spanning both the inner nuclear membrane (INM) and ONM and both envelopes are fused at these pores. These pores allow material transfer between compartments but also delineate the boundary between the two membranes which means they comprise distinct populations of integral and associated proteins (similar to a tight junction separating basolateral and apical sides of a cell). To give nuclei their rigidity they maintain their own nuclear cytoskeleton which connects, through multiprotein anchoring complexes spanning the NE (Nesprins and SUN proteins), with the cytoplasmic cytoskeleton (Figure 4A). This not only holds the nucleus in place but also links mechanical signals such as movement with the nuclear environment. On the nuclear side directly underlying the INM extends a multi-protein meshwork, the nuclear lamina, which is mainly made up of the intermediate filaments A- and B-type lamins. This filamentous layer has been implicated in several different functions including spacing of NPCs via interaction with Nup153, genome organisation, regulation of gene expression by acting as a repository for transcription factors and DNA replication. This may not come as a surprise as extensive connections between the lamina and chromatin form in interphase nuclei and its interruption manifests as pathology in certain genetic conditions (genetic laminopathies). A-type lamins include two alternatively spliced isoforms, lamin A and C, while there are two lamin B genes coding for two major and one minor lamin B protein. B-type lamins are farnesylated and methylated at a C-terminal CaaX motif (C- cysteine, a – aliphatic residue, X – any residue) which enables the insertion into the INM. A-type lamins on the other hand lose their farnesyl appendage through proteolytic cleavage of their C-terminal but still assemble at the nuclear periphery. Whether both lamin types interact at the nuclear rim is unknown but evidence suggests they can bind to one another in vitro. To maintain tight association of the meshwork with the INM other integral membrane proteins, such as lamin-B receptor (LBR), lamin-associated proteins (LAP1 and LAP2) and emerin, act as anchor points. Of these LBR, emerin and LAP2 isoforms bind lamin B while LAP1’s and emerin bind lamin A/C. Additionally, emerin, LAP2s and MAN1 (Inner nuclear membrane protein Man1) comprise a LEM (LAP2, emerin, MAN1) domain that serves as a chromatin anchor by recruiting the small DNA binding protein BAF (barrier to autointegration) and LBR separately recruits HP1 (heterochromatin protein 1) which thus indirectly links the lamina to DNA. This not only serves the structural organisation of nuclei but links the position of certain genetic areas to genetic silencing. An additional feature of proteins like emerin, LAP2s and MAN1 (and others) is that can also bind transcription factors activated by, amongst others, transforming growth factor (TGF)-β and Wnt,

- 30 -

Introduction mostly inhibiting their activity. Thus the lamina and nuclear structure directly feed back into mitotic signalling cascades likely only allowing signalling to occur when the conditions are right. The lamina is highly networked which makes it very difficult to extract these proteins directly and requires ionising detergents such as sodium dodecyl sulphate (SDS), but during mitosis rapid disintegration of the whole NE is observed. The process is kick-started with the simultaneous disassembly of all NPCs which coincides with a loss of nuclear transport and the build-up of cytoskeletal mechanical stress on the NE. Nuclear influx of kinases or their activating factors causes major destabilisation of the lamins which become extensively phosphorylated by kinases including cyclin-dependent kinase (CDK) 1 and protein kinase C (PKC) βII. The result of these processes leads to rapid mitotic NE breakdown (NEBD) which is a misleading term as the envelope is rather assimilated into the ER from where it is later re-assembled. Despite NEBD the expected complete loss of compartmentalisation is not observed during mitosis courtesy of several factors remaining associated with the segregating chromosomes (e.g. lamins and Ran guanine nucleotide exchange factor (RanGEF)). Furthermore, nucleoporin complexes released from the NE, in conjunction with importin- β, are re-purposed to regulate the ensuing chromosome segregation events by accumulating at the spindle poles. This juxtaposes them near the DNA when chromosomes reach the poles and thus facilitates appropriate re-assembly of the NE around the genetic material after CDK activity declines and lamin phosphorylation events are reversed by phosphatases including protein phosphatase 1 (information in the above paragraphs had been collectively reviewed in Dechat et al. 2008; Güttinger et al. 2009; Foisner 2001; Prokocimer et al. 2006; Wilson & Berk 2010; Brachner & Foisner 2011).

- 31 -

Introduction

Figure 4 Schematic representation of the NE and NPCs of metazoan cells. (A) Structure of the NE and associated proteins. (B) NPC showing all known Nucleoproteins and their approximate position within the NPC. Colour coding shows areas of the same composition. Images taken from (Güttinger et al. 2009).

The pores in the NE allow exchange of molecules between the nucleus and the cytoplasm, however, not all molecules enter the nucleus and trafficking is far faster than would be anticipated by simple diffusion. To maintain the integrity of the pore and allow selective gating they are fortified with a supramolecular complex made up of over 30 different proteins that weigh in at between 50-125MDa (Reichelt et al. 1990; Yang et al. 1998; Cronshaw et al. 2002; Stoffler et al. 1999). All its constituents are symmetrically arranged with an eight-fold axis (if viewed from the top) and each pore comprises four major bands (Figure 4B) which show two-fold symmetry (if viewed from the side). The two inner (magenta, red boxes) and outer (green, blue, brown, orange boxes) two rings, which are anchored by integral INM proteins POM121 (Pore membrane protein of 121 kDa) and gp210 (pore membrane protein of 210 kDa), are made up of the same complement of nucleoproteins (Nups), that is there is also a lot of symmetry between the two sides of the NPC. The inner rings of both sides, consisting of Nup205, 188, 93 and 155, anchor further Nups that emanate fibrils into the middle of the pore. These fibers contain hydrophobic dipeptide repeats of phenylalanine and glycine (FG repeats) which allow docking of transport substrates. Moving outward from the centre of the pore lies the outer ring, which comprises besides others the Nup107 complex. This complex is the platform for further FG-Nups which project their hydrophobic fibrils approximately 50nm into the cytoplasm. It is these mesh like appendages that obstruct free trafficking through the NPC and which allow directional movement by virtue of interactions with their FG repeats to the centre of the pore. To contribute to directional movement across the centre of the NE the symmetry has to be broken to prevent cargoes moving back to the cytoplasmic side. While the Nups themselves do allow cargoes to move in both directions it is

- 32 -

Introduction the nuclear transport factors anchored either in a gradient across the pore or at strategic positions within it. For instance, while the make-up of the outer nuclear ring is similar to the cytoplasmic side, the fibrils emanating into the nucleoplasm comprises different FG-Nups (Nup153-Tpr complex and others) which form a basket-like assembly. The basket harbours factors that are important for nuclear export of proteins, release of imported cargo proteins and for recycling of nuclear transport receptors (information in this paragraph was collectively reviewed in Dechat et al. 2008; Güttinger et al. 2009; Foisner 2001; Prokocimer et al. 2006; Wilson & Berk 2010; Brachner & Foisner 2011).

1.3.2 NLSs and nuclear import receptors

To facilitate the directed and fast movement of cargoes across the NPC a set of structurally related proteins evolved to cater for cargoes that range across the whole spectrum of protein sizes. These nuclear transport receptors are collectively termed karyopherins and include the alpha and beta sub-families which enable most transport processes. Karyopherins recognise their cargos through a diverse set of NLSs (NLSs) which either specify the classical or non-classical nuclear import pathway. Classical NLSs (cNLSs) have two defined aspects; they use the importin-α/β pathway to enter the nucleus and can roughly be divided into two classes, the mono- (mpNLS) and bi-partite (bpNLS) NLSs, based on the number of positively charged activity-conferring regions which are exemplified by the NLSs of the Simian Virus 40 (SV40) protein large T antigen (lTag) (Kalderon et al. 1984) and the cellular protein nucleoplasmin (Dingwall et al. 1982). Within the mpNLSs up to five different sub-classes have been defined which all comprise at least two out of five basic residues, although stronger signals feature at least three basics including an N-terminal lysine. The consensus is not absolute, that is, sequences may be extended and large hydrophobic, acidic or polar residues can be accommodated (Kosugi, Hasebe, Matsumura, et al. 2009). Despite this variety, most putative NLSs found by in silico analysis of the human genome fall into class 1 (Table 1). The much longer bpNLSs were suggested to consist of two imperfect mpNLSs which lie around 9-12 residues apart. However, sequences with larger linkers have been defined (Vasicova et al. 2013). While small proteins might not require an NLS to get to the nucleus their transport is considerably faster if one is present. Transport kinetics vary between cargoes and the rate of transport has been suggested to be proportional to relative affinity of the NLS to the transport receptor. This view, however, since it was based on experimental data of isolated NLSs or NLSs fused to GFP (Kosugi, Hasebe, Matsumura, et al. 2009; Hodel et al. 2006), is a little simplistic and transport rate will certainly be affected by the molecular environment (i.e. position of the NLS within cargo protein, size and structure of the cargo). Despite much research defining the sequence specific import modes the list of atypical NLSs is growing. Many of these proteins show no signal sequence that would be recognised by an NLS finder

- 33 -

Introduction algorithm (described in results section) but still end up in the nucleus. The first to be described was the glycine-rich NLS of hnRNP A1(heterogeneous nuclear ribonucleoprotein A1) (Siomi & Dreyfuss 1995) which enables binding to importin-β2 (transportin) (Aitchison et al. 1996). Most of these proteins bind to importin-β members directly rather than via importin-α and form multiple low affinity (high avidity) interactions with the concave surface of these proteins.

Table 1 Overview of the consensus sequences of the six classes of classical NLSs found in eukaryotic, multicellular systems. X = any amino acid; subscript shows number of a particular residue. The third column shows the absolute numbers of nuclear proteins (as annotated in uniprot) in H. sapiens that contain an NLS of this type. The fourth column shows the theoretical number of proteins in the human proteome (total sample size = 43571) containing a sequence matching the given consensus (Kosugi, Hasebe, Matsumura, et al. 2009; Chang et al. 2014).

No. of confirmed calculated no. of NLS class Consensus proteins proteins

Class 1 KR(K/R)R, K(K/R)RK 2473 16038 Class 2 (P/R)XXKR(D/E)(K/R) 1396 1707 Class 3 KRX(W/F/Y)XXAF 13 79 Class 4 (R/P)XXKR(K/R)(D/E) 1528 640 Class 5 LGKR(K/R)(W/F/Y) 4 25

Bipartite KRX10–12K(K/R)(K/R), KRX10– 998 580 12K(K/R)X(K/R)

In the last two decades it was brought to the field’s attention that the widely accepted NLS paradigm needs to be expanded. Nuclear transport as any other process is subject to extensive regulatory networks which include epitope masking and post-translational modification such as phosphorylation, acetylation and ubiquitination. This also raises the possibility that some NLSs might escape discovery as they are only active under certain circumstances. The karyopherin-β family is an ancient family spanning all eukaryotic lineages. Significant developmental steps can be correlated nicely with the appearance of new sub-groups of β- karyopherins (O’Reilly et al. 2011). The human β-karyopherin group comprises at least 18 members ranging in size from 876-1204 amino acids. These include both import (importin) and export (exportin) carrier proteins. While they only share significant sequence homology in the N-terminal RanGTP binding domain the rest of the protein remains structurally highly similar (see Figure 5B for importin- β1 structure) and each of the proteins shows a negatively charged overall character (probably to bind basically charged cargoes). Each member contains between 18-24 HEAT repeats which are short tandemly arranged modules of approximately 40 residues each consisting of an anti-parallel α-helix hairpin connected via a loop. These HEAT repeats are connected via linker regions and stacked with a slight clock-wise twist. Structurally they resemble a bent, superhelical rod comprising one concave and one convex surface (Marfori et al. 2011; Zachariae & Grubmüller 2008; Conti et al. 2006; O’Reilly et al.

- 34 -

Introduction

2011). This setup allows for extreme structural changes (Kappel et al. 2010) and by virtue of its flexibility can accommodate a variety of different cargoes which can be released extremely fast(Lee et al. 2005). While RanGTP (N-terminal), importin-α (C-terminal) and directly interacting cargoes (C- terminal) bind the concave surface of importin-β1, nucleoporins bind the convex surface via hydrophobic FxFG motifs. Binding to the convex surface appears to not affect binding of cargo while binding of RanGTP alters the overall structure enough to expel cargoes from the concave surface (importin-α or directly interacting proteins) in a spring-like manner. Importin-β1, at this time, seems to be the only β family member to bind to additional adapters including karyopherin αs, snurportin and hRIP isoforms, however, it appears possible that β karyopherins can form heterodimers to facilitate transport of specific cargoes as was found for importin-β1 and importin-7 mediated histon- shuttling (Jäkel et al. 1999). This implies that cells evolved more complex nuclear import pathways using more than one receptor at the time and implies that we will discover many more import pathways (or combinations of the existing ones) and quite likely find considerable redundancy at least in cell culture systems. The second group of major transport factors which are thought to contribute to at least 57% of all nuclear transport events in yeast (Lange et al. 2007) are karyopherin αs. These shorter transport proteins function as adapters between importin-β1 and cargo proteins, while most nuclear transport receptors, including importin-β1, interact directly with their cargoes. As this pathway was the first described to enable nuclear shuttling, it is termed the classical nuclear import pathway (Goldfarb et al. 2004). The appearance of karyopherin αs not only uncoupled evolution of importin-β1 cargoes from importin-β, which cannot change as much being facilitator of transport in many essential pathways, but it also increases delivery options and pools of cargoes (Lott & Cingolani 2011). It also allows cells to regulate nuclear localisation at a more intricate level and thus integrate diverse signals better (Riddick & Macara 2007). Thus far seven α family members have been discovered in humans which bind many different, at times overlapping cargoes. The seven members (importin-α1 through α7 or karyopherin α1-α8 while there is no α2) can be divided into three groups classed according to sequence conservation (α1s, α2s and α3s) (Goldfarb et al. 2004). The different α isoforms are tightly regulated and thus fulfil annotated roles during both cellular differentiation and cell proliferation (Koehler et al. 2002; Pulliam et al. 2009). All eukaryotic cells possess homologues of the α1s while multicellular organisms such as humans contain all three classes enabling more complex regulation through tissue and time specific expression. In vivo studies in animals that comprise one or few members of each of the three classes (Drosophila and C. elegans) yielded valuable functional information. As such, α2s were dispensable during development but fruit flies were sterile and C. elegans produced non-viable

- 35 -

Introduction offspring if the factor was deleted while knock-out of α3 showed additional defects in cell growth on a more general level (Goldfarb et al. 2004).

Figure 5 Structural representation of importin proteins. Importin-α (A) and importin-β (B) are depicted as ribbon structure showing Armadillo (α) and HEAT (β) repeat arrangements. Image taken from (Goldfarb et al. 2004).

Structurally, importin-αs are smaller than the β family members and weigh in at approximately 50kDa and ranging in size from 499 to 538 residues (UniprotKB database). All feature a short N-terminal auto-inhibitory domain (Figure 5B, importin-β binding domain, IBB, bound to importin-β), which also facilitates binding to importin-β1, and a C-terminal cargo binding domain made up of ten armadillo (ARM) repeats, which consist of three α-helices (Figure 5A). The Armadillo (ARM) repeats structurally resemble the HEAT (Huntingtin, elongation factor 3, protein phosphatase 2A, and the yeast kinase TOR1) repeats and result in similar regulatory features. For instance, the N-terminal IBB domain when released by importin-β1 and the tenth ARM repeat, which is bound by CAS (cellular apoptosis susceptibility) and/or Nup50, facilitate efficient cargo release due to conformational changes in importin-α (Gilchrist et al. 2002; Matsuura et al. 2003; Goldfarb et al. 2004). Similarly, the ARM repeats allow a high degree of flexibility to accommodate cargo proteins and different NLSs (Roman et al. 2013). Binding of cargoes is facilitated by the two NLS binding sites, the major groove and the minor groove, formed by ARM2-4 and ARM7-9, respectively. While mono-partite NLSs can bind to the major or minor groove, bi-partite usually make contact with both simultaneously and the distance of both grooves sets the optimum spacer length of bi-partite NLSs to between 10-12 residues.

1.3.3 RanGTP cycle and mechanism of nuclear import

Nuclear import is a very complex process and it becomes increasingly obvious that the classical paradigm of nuclear transport has many exceptions. Over time the ‘diffusion limit’ of the nuclear pore

- 36 -

Introduction has been based on easily purified substrates (bovine serum albumin (BSA) and ovalbumin) and was raised sequentially (Paine et al. 1975; Keminer & Peters 1999; Görlich & Kutay 1999), although proteins as large as 110 kDa (Wang & Brattain 2007) have been found to enter nuclei of cells despite the absence of any discernible NLS or requirement for ATP (indicating active transport). Diffusion through the pore is restricted by unfolded nucleoporins that emanate into the central channel which sets the navigable space to approximately 9 nm (Paine et al. 1975; Denning et al. 2003). It has now become widely accepted that molecular weight is not the only determining feature since globular proteins might be obstructed from entering the nucleus due to exceeding the estimated diameter of the NPC of 9 nm and sometimes smaller proteins were found to diffuse slower than larger ones (Görlich & Kutay 1999). In the presence of a targeting signal, however, the diffusion barrier is less of a problem and theoretically only the full diameter of the NPC can limit import of proteins, although gold particles as large as 39 nm were shown to pass the NE in the presence of an NLS (Panté & Kann 2002; Dworetzky et al. 1988). Even though diffusion is possible for very small proteins it may be of importance to allow fast entry or it may be vital that they are exclusively nuclear (e.g. histone family) and thus targeting signals are widely distributed throughout the proteome (see Table 1 for distribution of classical NLSs). For active nuclear transport to be sustained a protein requires besides an NLS: (i) the ability to dock at the NPC directly or via nuclear transport adapters (ii) asymmetrical distribution of some nucleoporins and, most importantly, RanGTP to enable directional transport of cargoes (iii) energy in the form of ATP/GTP (iv) a factor to enable its release from the nuclear transport machinery in the nucleoplasm The process of nuclear import, here illustrated using the paradigmatic classical transport pathway (for examples of cargoes of the different transporters see Harel & Forbes 2004; Pemberton & Paschal 2005; Sorokin et al. 2007), begins in the cytoplasm. Cargo proteins are recognised by importin-α isoforms by virtue of their basic NLS. It is generally thought that NLSs form linear epitopes binding to the concave surface of importin-α at one of its two binding grooves. In its free form importin-α exists in an auto-inhibited form with its N-terminal IBB domain blocking the NLS binding sites. Two not mutually exclusive scenarios were suggested to regulate importin-α binding activity. Either importin-β binds to the IBB, which increases the affinity to cargo NLSs substantially or NLSs displace the IBB which allows binding to importin-β. It would perhaps not be surprising if both processes work simultaneously to allow optimum binding of cargoes. While the majority of mpNLSs binds the major groove (classes 1 and 2), a smaller proportion binds solely to the minor groove, and bpNLSs appear to bind both simultaneously. Formation of the ternary complex (importin-α/β-cargo) allows binding to NPC via the convex surface of importin-β. The complex then moves hand over hand

- 37 -

Introduction along the FxFG repeat ‘gradient’ through the channel as some of these hydrophobic repeat proteins are asymmetrically distributed. Inhibition of FxFG protein binding, for example through wheat germ agglutinin, which binds these proteins, abolishes directed nuclear transport. This process relies on on- off rates for FxFG/importin-β interactions and likely on the presence of two separate Nup binding sites in importin-β (Harel & Forbes 2004; Bednenko et al. 2003; Kutay et al. 1997). Once the ternary complex reaches the nuclear basket, release of the cargo proteins is facilitated by binding of RanGTP which both sterically displaces it and causes allosteric changes incompatible with importin-α/cargo binding (Cingolani et al. 1999). Importin-β is subsequently recycled back into the cytoplasm in complex with RanGTP which prevents it from binding further cargo proteins. In the cytoplasm Ran hydrolyses GTP to GDP, stimulated by the activity of the RanGAP1 (Ran GTPase activating protein 1) complex and RanGTP binding to Nup358 (Ran binding protein 2 (RanBP2)), causing dissociation of importin-β (Yokoyama et al. 1995; Bischoff & Görlich 1997). The nuclear importin-α/cargo complex dissociates independently in a process that likely commences through importin-β release of the auto-inhibitory IBB domain and is assisted by binding of the CAS/RanGTP complex, the designated recycling transport factor of importin- α, and Nup50 to the C-terminus. CAS export works similarly to Crm1 export in that it facilitates reverse transit by moving along Nups and releases its cargo upon conversion of GTP to GDP, which is stimulated by cytoplasmic RanGAP1. Pivotal to directional transport across the NPC is the strategic placement of certain regulator proteins within the cell (reviewed in (Kalab & Heald 2008)). The determining factor is the asymmetrical distribution of both nucleotide-bound forms of Ran (GDP and GTP) which is maintained by exclusively cytoplasmically localised RanGAP1 and chromosome bound RanGEF RCC1 (Regulator of chromosome condensation 1). When importin-β complexes are dissociated in the nuclear basket by RanGTP this complex migrates to the cytoplasmic side where NPC-localised RanBP1 and RanGAP1 stimulate the Ran-GTPase activity (Bischoff & Görlich 1997; Mahajan et al. 1997). This facilitates dissociation from importin-β but in turn would cause cytoplasmic accumulation of RanGDP. However, in its GDP-bound form Ran is recycled back to the nucleus by NTF2 (nuclear transport factor 2). Unlike most importin-β mediated transport events, NTF2 dimerises to mediate movement through the pore by sequentially binding the relevant Nups (Bayliss et al. 2002). This one-way journey completes through the RCC1 mediated nucleotide exchange which releases RanGTP from the complex. NTF2 in turn ‘exports itself’ upon release of its cargo to close the cycle.

- 38 -

Introduction

1.4 Viral exploitation and modulation of nucleo-cytoplasmic transport

Many viruses besides the HVs, including most DNA virus (except pox viruses) and even some RNA viruses (e.g. human immunodeficiency virus (HIV)) must enter the nucleus to complete their replicative cycle. However, there are several obstacles to reaching their destination. Firstly, they have to traverse the cytoplasmic membrane which, depending on the virus in question, can be via several routes; (i) Receptor mediated membrane fusion for enveloped viruses as found for, e.g. retroviruses(Gallo et al. 2003) and HVs(Browne et al. 2001) (ii) Receptor mediated endocytosis (Hasebe et al. 2009; Mercer & Helenius 2009; Grove & Marsh 2011), including (macro-)pinocytosis (Nicola et al. 2005; Saeed et al. 2010), caveolin-dependent (Hasebe et al. 2009; Anderson et al. 1996) uptake and clarthrin- mediated endocytosis (Matlin et al. 1981), of both enveloped and non-enveloped viruses, e.g. , HVs (in some cell types) and influenza (iii) Disruption of the cytoplasmic membrane by non-enveloped viruses (Hogle 2002)

To add to the complexity of virus entry to cells, many viruses were shown to utilise different pathways in different cell types. It is conceivable that viruses entering their target cells via multiple pathways do this by default and, depending on the physiology of the target cell, reach their destination by the one that works best. This would be beneficial to an entering virus since the body is highly differentiated ensemble of cells that harbour a very specialised physiology and have at their disposal several potential anti-viral mechanisms. Quite often viruses were found to end up in dead-end pathways (Gershey & Diacumakos 1978; Diacumakos & Gershey 1977; Thomas et al. 2007; Grove & Marsh 2011) leading to their recognition and destruction and depending on the cell type only some would lead to infection. For instance, only approximately 1% of SV40 polyomavirus particles, for example, were shown to cause an infection (Diacumakos & Gershey 1977; Gershey & Diacumakos 1978) when virus is added to cells, which can be increased by 1 log unit each by injecting virus into the cytosol or nucleus, implying that viral entry is a race which is governed by stochastical events. Additionally, observations from HV infection in our lab indicate that the same HSV-1 inoculum yields different numbers of plaques depending on the cell type infected. Notwithstanding the various different entry mechanisms viruses employ, virus capsids/particles have to traverse the densely packed cytosol. This can achieved either by directly/indirectly engaging the host cytoskeletal machinery via capsid components or by deploying the endosomal sorting components as adapters. Prior to nuclear entry virus particles/capsids usually have to enter the

- 39 -

Introduction cytosol, except for some described exceptions (Nishimura et al. 1991; Butin-Israeli et al. 2010; Butin- Israeli et al. 2011), via endosomal lysis or fusion to be able to penetrate the nucleus (Grove & Marsh 2011). In the last three decades a few routes into the nuclear compartment have been described. These include capsid movement through the nuclear pore (Baculovirus), capsid docking at the NPC (HSV) which may involve capsid disassembly (Adenovirus), disruption of the NE (parvoviruses), fusion with the INM (SV40), transport of previously uncoated ribonucleoproteins (RNPs) through the NPC (Influenza and HIV) or NEBD (non-lenti retroviruses). The following paragraphs will summarise these pathways using representative examples.

- 40 -

Introduction

Table 2 Overview of how selected viruses use and/or modify the cellular nuclear transport/pore machinery or the NE itself to enter cells.

Group Family Virus Viral Component Targets Process Reference capsid Importin-β capsid docking at nuclear pore (Ojala et al. 2000) capsid Nup358 capsid docking at nuclear pore (Copeland et al. 2009)

(Copeland et al. 2009; pUL25, capsid Nup214 capsid docking at nuclear pore; DNA release Pasdeloup et al. 2009) HSV-1 binds free pUL25; might be involved in capsid pUL25 hCG1 (NUPL-2) (Pasdeloup et al. 2009) docking at nuclear pore

Herpesviridae uncoating of DNA at the NPC; soluble pUL25 pUL25 ? likely NPC component (Rode et al. 2011) inhibited uncoating but not NPC docking (Nakanishi et al. 2002; capsid delivery to the nucleus; all VPs contain Nakanishi et al. 2007; Clever general nuclear transport VP1/2/3 classical NLSs which likely use the importin-α/β et al. 1991; Yamada & machinery pathway Kasamatsu 1993; Nakanishi et al. 1996) SV40 disruption of nuclear lamins by caspase-6 in order for capsids to penetrate INM from ER; fluctuations

Polyomaviridae VP1 pentamer (Butin-Israeli et al. 2010; caspase-6, Lamin A/C in Lamin A/C levels and associated (unknown mechanism) Butin-Israeli et al. 2011) 1 phosphorylation at an unidentified residue in

Entry quiescent cells (Saphire et al. 2000; Strunze Ad2/5 Crm1, Hsp70 nuclear targeting of capsids et al. 2005) Nup214, Importin-β and nuclear targeting of capsids via adapter protein hexon (Trotman et al. 2001)

importin-7, Histone H1 histone H1 capsids bind Nup214/88/358 to dock at the NPC; Nup358 recruits kinesin heavy chain Kif5c to hexon Nup214/62/358 (Strunze et al. 2011) Ad2 promote capsid disintegration; simultaneously

Adenoviridae leads to NPC perturbation Importin-α/β, Importin-7, (Wodrich et al. 2006; Hindley pVII nuclear import of Ad genomic DNA post-docking Transportin-1 et al. 2007) hexon Importin-α/β, Nups nuclear docking of capsids (Saphire et al. 2000)

Baculo- (Ohkawa et al. 2010; van Loo AcMNPV Importin-β nuclear delivery of capsids viridae et al. 2001)

- 41 -

Introduction

Group Family Virus Viral Component Targets Process Reference proteasome nuclear delivery of capsids Ros 2004 MVM (Cohen & Panté 2005; Cohen Caspase-3, lamin B2 NE disruption (ONM) by digestion of lamin B2 et al. 2006; Cohen et al. 2011)

? (unknown mechanism but independent from wtAAV2 ? (Hansen et al. 2001) trafficking through the NPC)

viridae

- recAAV2 Importin-β, Importin-7, NPC nuclear delivery of capsids (Nicolson & Samulski 2014)

Parvo binding to NPC via alternative mechanism leads to Nup62, Nup153, Nup214, subsequent disruption of the ONM and INM via H1 Nup358, PKCα, Cdk1/2, (Porwal et al. 2013) the mitotic kinase-protease axis Caspase-3, Lamins

(Wang et al. 1997; Cros et al. NP Importin-α nuclear import of vRNP 2005; Wu et al. 2007; Melen et al. 2003; Ozawa et al. 2007) Importin-α1 or 5 + Importin- NP nuclear import of vRNP (O’Neill et al. 1995) β

NP Transportin-3 nuclear import of vRNP (RNAi) (König et al. 2010) CSE1L, KPNB 1, 5 potentially nuclear import of vRNP (RNAi) (König et al. 2010) NUP214,NUP153, TNPO3

Influenza A Influenza Importin-β, Crm1, Nup98, potentially nuclear import of vRNP (RNAi) (Karlas et al. 2010)

Orthomyxoviridae Nup205 likely nuclear import of vRNP as delayed gene Importin-5 (Deng et al. 2006) expression PB1/2, NP, PA + NS1 Importin-α4 (Shapira et al. 2009) PB1/2, NP, PA Importin-α7 (Shapira et al. 2009) NEBD if Vpr overexpressed; greatly improves (Vodicka et al. 1998; Popov et docking of the PIC at the NPC by directly binding to al. 1998; de Noronha et al. Vpr hCG1, Nup54, Importin-α/β nucleoporins; increased affinity of importin-α/β to

2001; Jacquot et al. 2007; Le the PIC; Vpr at the NE promotes G2 Rouzic et al. 2002) arrest/apoptosis 6 HIV-1 chromatin associated Nup62 to aid genome Integrase Nup62 (Ao et al. 2012) integration

Retroviridae (Zaitseva et al. 2009; Ao et al. Integrase Importin-7 nuclear transport of pre-integration complex (PIC) 2007; Fassati et al. 2003) Integrase Nup98 reduction of genome integration (Ebina et al. 2004)

- 42 -

Introduction

Group Family Virus Viral Component Targets Process Reference Integrase Importin-α1 nuclear import of PIC (Gallay et al. 1997) Integrase Importin-α3 nuclear import of PIC into T cell nuclei (Ao et al. 2010) (Krishnan et al. 2010; Christ et Integrase, CA Transportin-3 nuclear import of PIC al. 2008) CA Nup153, Nup155 nuclear import of PIC (Lee et al. 2010) targeting of imported PIC and increased efficiency CA Nup358 (Schaller et al. 2011) of integration in transcriptionally active areas (Haffar et al. 2000; Bukrinsky Matrix nuclear import of PIC into non-dividing cells et al. 1993) masking of receptor binding MLV CA protein inhibits nuclear transport of PIC (Yamashita & Emerman 2004) site

- capsid entry to the nucleus and capsid arrest at (Kann et al. 1999; Schmitz et 7 HBV core protein importin α/β, Nup153 Nup153 acts as a maturity checkpoint (coupled al. 2010) viridae potentially to phosphorylation state)

Hepadna

- 43 -

Introduction

1.4.1 Adenovirus

Adenoviruses are large non-enveloped DNA viruses that infect cells through specific interactions of the fibre knob domain to the CAR (Coxsackie Adenovirus receptor) or CD46 (cluster of differentiation 46) viral receptors, which leads to clathrin-dependent endocytosis. This initial attachment is facilitated by the fibre knob domain. It is within these pre-early endosomes that structural pH-dependent (Maier et al. 2012; Seth et al. 1984; Seth 1994) changes occur liberating pVI (Adenovirus protein 6) from the virion leading to lysis of the endosome (Prchla et al. 1995). Cytosolic capsids migrate towards the nucleus along microtubules which requires hexon dependent recruitment of dynein motor proteins (Smith et al. 2008). The large adenovirus capsid diameter of approximately 90nm cannot enter the nucleus through the NPC and a complex series of disassembly events occurs at the NE which leads to release of the viral DNA and subsequent transport through the central channel of the pore. Initial evidence implies that nuclear pores or at least docking there, are required for disassembly of capsids and subsequent import of viral nucleic acid into the nucleus. Intact capsids can easily be seen at the NE (Greber et al. 1996). When wheat germ agglutinin or anti-NPC antibody (clone RL1), both inhibitors of nucleo-cytoplasmic transport (Akey & Goldfarb 1989; Finlay et al. 1987; Featherstone et al. 1988), were injected into cells adenovirus disassembly was blocked and free penton or nuclear pVII could not be detected. That implied contact with the NPC is required to allow disassembly (Greber et al. 1997; Snow et al. 1987). For actual attachment to the pore hexon and facilitates capsid binding to Nup62, 214 and 358, the latter recruiting kinesin heavy chain Kif5c. The close proximity to the nuclear environment also allows hexon protein to recruit histone H1, which, in conjunction with Kinesin, in turn was found to initiate capsid disassembly (Strunze et al. 2011; Trotman et al. 2001; Saphire et al. 2000). Surprisingly, Crm1 (chromosome maintenance 1 or exportin-1), the bona fide nuclear export protein, contributed an essential but unknown function to the entry process which likely involves release of capsids from microtubules upon reaching the NPC. After uncoating the DNA- pVII complex recruits nuclear transport factors Importin-α/β to initiate transport of the viral genetic material to the nucleoplasm, a process which also requires the protein Hsp70 (heat shock protein 70) (Saphire et al. 2000).

1.4.2 HIV

Retroviruses are positive sense single-stranded RNA (ssRNA) viruses (Baltimore group 6) that integrate their genetic material into host cell chromatin. Cryptic remnants can be found all over the human genome indicating a long-standing relationship. All retroviruses, usually in form of a pre- integration complex, require access to the nucleus. While murine leukaemia virus (MLV), a γ-retrovirus,

- 44 -

Introduction gains access to the nucleus through mitotic NEBD (Roe et al. 1993), HIV, a lentivirus, can also integrate into chromatin of quiescent cells (Gartner et al. 1986). After receptor binding (CD4 and C-C chemokine receptor 5 (CCR5) or C-X-C chemokine receptor 4 (CXCR4)) by gp120 at the cytoplasmic membrane fusion and deposition of viral capsid into the cytoplasm occurs. Nuclear entry requires partial uncoating and loss of some capsid (CA) protein to take place. Recruitment of cellular factors allows migration along microtubules towards the nuclear periphery. There is some debate about whether reverse transcription occurs during transit or at the nuclear pore. Nonetheless the viral pre-integration complex (PIC) needs to bind to the NPC. In the last three decades since its discovery almost every PIC component, including reverse transcription products, have been implicated, by virtue of their karyophilic nature and ability to bind NPC components, to be important for nuclear import of the HIV genome (reviewed in Matreyek & Engelman 2013). Both canonical (matrix protein (MA), integrase (IN), viral protein R (Vpr)) and non-canonical (Vpr, CA) NLSs were shown to be functional for soluble protein import, recruiting a diverse set of nuclear transport factors which include importin-α1 and 3, importin-β1, importin-7 and transportin-3 (Table 2; also for references). However, subsequent studies also indicated that deletion of karyophilic signals in each of the PIC proteins still allowed infection of quiescent cells adding inconvenient complexity to the enigmatic and thus far pharmaceutically untouched process of nuclear import of PICs. As for many other viruses, these processes are likely redundant and dependent on the physiological state of the target cell. It is also conceivable that multiple importins are required for distinct steps of PIC nuclear import and integration. As such, one or more factors may be required for import per se while others could guide or protect the PIC (Valle-Casuso et al. 2012; Levin et al. 2010; Fricke et al. 2013; De Iaco et al. 2013; Ao et al. 2012; Schaller et al. 2011). Nucleoporins also play an important role during nuclear delivery of PICs. As such, Vpr, IN and CA have all been shown to interact with Nup54, 62, 98, 153, 155, 358 and nucleoporin-like protein 1 (hCG1) (Table 2; including references). These likely allow importin independent docking at the NPC. In fact, the Vpr-hCG1, Vpr-Nup54, IN-Nup98 and CA-Nup153/155 interactions can anchor the PIC to the nuclear pore while this likely increases affinity to resident nuclear transport factors such as importin- α/β. Interestingly, a HIV-1 chimera carrying MLV-CA protein could only infect cycling cells. The block here, however, might relate to integration rather than PIC transport (Yamashita & Emerman 2004). Recent findings add another layer of complexity since several Nups can bind to chromatin structures and modulate diverse processes including transcription and gene silencing (Ruben et al. 2011; Van De Vosse et al. 2013; Kehat et al. 2011; Capelson et al. 2010; Kalverda et al. 2010). It appears these could

- 45 -

Introduction be used, in addition to entry, to target the viral PIC to specific integration sites post-nuclear entry (Ao et al. 2012; Schaller et al. 2011).

1.4.3 Influenza

The Orthomyxoviruses, exemplified by Influenza A virus (IAV), are nuclearly replicating, segmented negative sense ssRNA viruses that have to enable each RNP complex to shuttle to the nucleus independently. Infecting virions enter non-ciliated cells via terminal sialic acid moieties of glycoproteins which are recognised by the Hemagglutinin (HA) virion surface antigen (Knipe & Howley 2013). This leads to endocytosis and subsequent uncoating (Chen & Zhuang 2008; Eierhoff et al. 2010). Alternatively, Influenza may be internalised by macropinocytosis (Rossman et al. 2012). In both cases release into the cytosol requires an acidic pH within the virion, which, in the case of influenza A, is facilitated by the H+ channel M2 located on the virion envelope (reviewed in (Pinto & Lamb 2006)). In late endosomes the low ambient pH induces conformational changes to matrix protein 2 (M2) which in turn leads to acidification of the virion interior which causes additional conformational changes to release M1 from RNPs (Bui et al. 1996) and to enable HA-mediated fusion ((Bullough et al. 1994) and reviewed in (Skehel & Wiley 2000)). The liberated RNPs, consisting of genomic RNA wrapped in nucleoprotein (NP) and capped by the viral polymerase (PB1/2 and PA), rapidly migrate towards the nucleus through unknown forces, although diffusion has been suggested (Martin & Helenius 1991; Chou et al. 2013; Babcock et al. 2004). There is a growing list of nuclear transport receptors including importin-α1, 2, 4, 7, importin-β1 and potentially importin-5, Transportin-3 and the nuclear export receptor Crm1, that have been implicated in facilitating infection with influenza (see Table 2 for references). NP was found to bind the classical complement of transport proteins which, as the main component of the RNPs, contributes significantly to nuclear import and thus infectivity. Although NP comprises three confirmed NLSs (Bullido et al. 2000; Neumann et al. 1997; Weber et al. 1998; Wang et al. 1997), but only the most N-terminal of which was found to be required for RNP nuclear import (Cros et al. 2005). In their hands peptides mimicking the unconventional NLS of NP solely blocked nuclear accumulation of RNPs. The advent of screening cells for cellular factors allowed further advances in delineating the proteins enabling infection. These screens found interactions of RNP components with importin-α4 and 7 and indicated that knockdown of importin-β1, Transportin-3 and Crm1 reduced infection in vitro (Karlas et al. 2010; König et al. 2010). As influenza is assembled in the nucleus this may not necessarily relate to effect on entry but on general replication and/or assembly. Crm1, for instance, shuttles RNPs, probably via interaction with NP, to the cytoplasm (Elton et al. 2001). Once the importin-RNP complex reaches the NPC, it makes contacts with individual nucleoporins (Babcock et al. 2004) although the

- 46 -

Introduction exact mechanism is unclear. Knockdown of individual Nups has been shown to reduce infection which may relate to entry of virus into cells (König et al. 2010; Karlas et al. 2010) but could equally impede replication later on. Data obtained by Babcock et al. shed light on some regulatory aspects of nuclear entry implicating M1 in inhibition of nuclear re-entry of newly synthesised RNPs (Babcock et al. 2004).

1.4.4 Baculovirus

Baculoviruses are large (30-60nm diameter) rod-shaped dsDNA viruses that infect insect and mammalian cells. Its glycoprotein Gp64 interacts with an as yet unknown receptor to induce clathrin- mediated endocytosis (Long et al. 2006; Hefferon et al. 1999). After approximately 30 min cytoplasmic nucleocapsids induce directional actin polymerisation by virtue of polarised distribution of VP78/83 to migrate towards the nucleus (Xu et al. 2007; Ohkawa et al. 2010). While early evidence suggested that Betabaculoviruses inject their DNA via polar attachment of capsids at the NPC (reviewed in (Au et al. 2013)) it is now widely accepted that Alphabaculovirus capsids may also enter the nucleus through another importin-β mediated mechanism (Loo et al. 2001; Au & Panté 2012). Using a rather unconventional approach, whole capsids entered the nucleus through the pore which meant that, since capsids can measure 30-60 nm in diameter, the NPC is more flexible than anticipated. In the case of baculoviruses the required NPC reorganisation was driven by a thus far unknown signal, although it was shown that nuclear accumulation of capsids was blocked by wheat germ agglutinin and truncated importin-β (Au & Panté 2012; Loo et al. 2001). This at least indicated that direct interactions with the NPC and the nuclear transport machinery are required. Further studies will show whether capsids directly or indirectly affect the constitution and shape of the NPC.

1.4.5 Viral subversion of nucleocytoplasmic transport

The NE does not merely function as a barrier to separate nuclear and cytoplasmic proteins but in combination with the many nucleo-cytoplasmic transport receptors has also been implicated in a growing list of processes including differentiation, inflammation, circadian rhythm, mitosis, gene expression and carcinogenesis. All viruses that infect nucleated cells in some way need to modulate nucleo-cytoplasmic transport. Besides exploiting the cellular nuclear transport machinery to facilitate entry of viral nucleic acid into the nucleus, viruses also regulate the import/export pathway to shuttle viral proteins, export mRNA, maintain/modulate cellular protein functions, facilitate assembly and to inhibit antiviral responses (see Table 3 including references). Most of the studies concerning viruses have focussed on entry and, albeit less frequently, on exit of viral particles as these represent the obvious anti-viral targets. Since the advent of genomic techniques evidence is mounting that viruses

- 47 -

Introduction may alter the composition of the NE substantially to change nucleo-cytoplasmic transport on a global scale. This may not be surprising as host gene expression is significantly affected and sometimes heavily impaired during many viral infections such as with HSV-1. HSV-1 infected cells produce a large amount of viral capsids which accumulate in nuclei late during infection. At a diameter of over 100nm they are far too large to exit the nucleus through the nuclear pore so must have evolved to use an alternative pathway which does not compromise nuclear integrity. In fact, through studying α-HVs, such as HSV-1 and PRV, a completely new nuclear exit pathway was discovered, which involves budding into the perinuclear space at the INM followed by fusion with the ONM (Darlington & Moss 1968; Speese et al. 2012). Two viral proteins, pUL31 and pUL34, in conjunction with additional cellular factors and the viral kinase pUS3, facilitate the INM budding/ONM fusion process. Intriguingly, in cells transfected with pUL31 and pUL34 (PRV) these form a complex at the INM and mediate budding of vesicles into the perinuclear space in the absence of infection (Fuchs, Klupp, et al. 2002). This process requires the activity of kinases including PKCα which phosphorylate A-type lamins and emerin causing a local disruption of the lamin mesh which underlies the NE (Table 3, also for references). It is at such locations where laminar thinning allows capsids to bud into the perinuclear space. This pathway may be considered a common theme amongst HVs as an analogous process is driven by the pUL50, pUL53 and pUL97 proteins of HCMV (and its mouse homologue). In fact, HCMV goes even further and reshapes nuclei to lobe around a central assembly compartment that is formed over the MTOC. It is at the concave surface of the NE very close to the assembly compartment that capsids exit the nucleus to enter the final steps of assembly. Besides the conventional egress pathway, HSV-1 and bovine HV 1, were shown to cause a reduction in the number of NPCs and, more strikingly, dilation of the remaining pores which could potentially enable capsid egress. Intriguingly, recently it emerged that in the absence of the nuclear egress complex (pUL31/34) compensatory (rescuing) mutations facilitate egress of both empty and filled capsids from the nucleus. It seems that through the concerted action of several viral proteins the cellular MEK kinase cascade is induced culminating in activation of roscovitine sensitive CDKs (cyclin dependent kinases). This in turn leads to NEBD by phosphorylation events similar to mitosis (Table 3) leading to release of A, B and C capsids. Wild-type HSV-1, probably as a capsid quality control mechanism, actively prevents mitosis, although it relies on CDK activity throughout its life cycle (Schang et al. 1998; Schang et al. 1999), and so must have evolved ways to stop wide spread breakdown of lamins even in the presence of phosphorylation. For instance, HSV-1 infection was found to cause inhibition of caspases 3 and 8 and dislocate CDK-5 which contribute to NEBD during apoptosis (Jerome

- 48 -

Introduction et al. 1999; Jerome et al. 2001; Mostafa et al. 2015) and deletion of pUS3 appeared to enhance NEBD (Maric et al. 2014). In contrast to HSV-1 other viruses modulate NPC function to alter cellular physiology and likely their nuclear assembly and egress. Recently evidence is mounting that HIV-1 infection alters NPC function and composition on a global scale (Table 3 including references). The protein level of numerous NPC components was reduced which likely affects nucleo-cytoplasmic transport and gears trafficking to favour viral production. On an additional albeit more subtle level, Rev protein (HIV) directly binds Crm1 to facilitate viral RNA (vRNA) export and Nup62 is redistributed to the cytoplasm, which indirectly affects Transportin-1 mediated nuclear import. The latter was also found to be incorporated into virions. Most RNA viruses are strongly dependent on their ability to antagonise the interferon pathway. Ebolavirus can infect humans and the fate of the infected very much depends upon how quickly an interferon response can be mounted. To counteract the antiviral signalling the virus encoded VP24 for example directly binds to importin alpha-5, 6 and 7 to inhibit STAT1 (signal transducer and activator of transcription 1) translocation to the nucleus which is a prerequisite of interferon production (Table 3 including references). The activity of VP24 thus likely contributes to the severe pathology encountered in this viral infection. Through the binding to multiple import factors many cargoes should be affected in a similar fashion but to date the only other protein found to be prevented from entering nuclei is hnRNP C1/C2. The examples given above are only a select set of which there is an extended list in Table 3. The list of viral proteins that modulate nucleo-cytoplasmic proteins continues to grow as we investigate viral protein functions and will likely yield valuable insights into cellular physiology and open up new opportunities to develop anti-virals.

- 49 -

Introduction

Table 3 Overview of how selected viruses subvert the nuclear transport machinery to facilitate efficient replication and virion egress.

Viral Group Family Virus Targets Process Reference Component lamin redistribution through recruitment of (Reynolds et al. 2002; Bjerke & Roller Emerin, lamins and lamin B PKC, Rottlerin sensitive kinase, pUS3 and 2006; Mou et al. 2007; Cano-Monreal pUL31/34, receptor, emerin, Rottlerin pUL13 by pUL31/34; phosphorylation of et al. 2009; Scott & O’Hare 2001; pUS3, pUL13 sensitive kinases, PKC (α, d) emerin and lamins; causes INM budding of Simpson-holley et al. 2004; Klupp et al. capsids 2007; Leach et al. 2007) pUL49, Roscovitine sensitive breakdown of nuclear pUL46, integrity during infection to allow exit of (Klupp et al. 2011; Grimm et al. 2012; pUL21, CDKs, MEK1/2 HSV-1 pUL31/UL34 virus mutant; may reduce Schulz et al. 2014) pUS2, gI, capsid quality control pUS9, pUL26 pUS3 mediated phosphorylation of gB; pUL31/34, (Farnsworth et al. 2007; Wisner et al. ONM fusion with ONM to release capsid into pUS3, gB, gH 2009; Wright et al. 2009) cytoplasm

UL34, gB, gH NE NE breakdown in TorsinA deficient cells (Maric et al. 2014) 1 Invagination of inner nuclear membrane; M50, M53, (Muranyi et al. 2002; Marschall et al.

Egress/NEBD p32,PKC, Lamins redistribution of lamins; increased surface Herpesviridae pUL97 2005; Buser et al. 2007) area; capsid egress M50/53 complex causes redistribution of PKC to INM; phosphorylation and HCMV M50/M53 PKC (Ca2+) (Muranyi et al. 2002) subsequent dispersal of at least lamins A and MCMV C

p32 recruits pUL97 to INM (specifically to pUL97 p32/Lamin B receptor, Lamin A/C LBR); pUL97 phosphorylates p32 and lamin B (Marschall et al. 2005) to redistribute lamins; capsid egress

dilated nuclear pores for capsid egress; HSV-1 decreased number of pore with increased ? NPC (Wild et al. 2009; Wild et al. 2005) boHV-1 signal of nucleoporin by immunofluorescence

- 50 -

Introduction

Viral Group Family Virus Targets Process Reference Component

(O’Neill et al. 1998; Neumann et al.

2000; Watanabe et al. 2001; Elton et al. NEP, M1 Crm1, RanGTP nuclear export of vRNP through NESs in NEP 2001; Shimizu et al. 2011; Huang et al. 5 2013)

Influenza A Influenza (Pleschka et al. 2001; Ludwig et al. HA PKCα and ERK pathway nuclear export of vRNP

Orthomyxoviridae 2004; Marjuki et al. 2006)

Through binding Nup62 nucleocytoplasmic ICP27 Nup62 transport via Importin α/β, Transportins are (Malik et al. 2012) inhibited

ICP27 NXF1 viral unspliced mRNA exported (Johnson et al. 2009) HSV-1 (Hofemeister & O’Hare 2008; Leuzinger Nup153 redistribution to the cytoplasm et al. 2005; Nagel et al. 2008) increased O-glycosylation likely at Nups; Nucleoporins, Nup358 (Hofemeister & O’Hare 2008) Herpesviridae reduction in complexed Nup358 binds Nups to phosphorylate and inhibit Importin-β mediated nuclear import and 1 EBV BGLF4 Nup62, Nup153 (Chang et al. 2015) promotes non-NLS mediated import of lytic proteins

capsid binding and disassembly causes NPC Ad2 Nup214/62/358 (Strunze et al. 2011) perturbation through displacement E1B55, mRNA processing machinery; mRNA export inhibition via unknown Ad type C (Yatherajam et al. 2011) E4Orf6? NXF1 mechanism

Adenoviridae

- inhibition of nuclear transport of transportin L1 Transportin-1 (Pollard et al. 1996; Nelson et al. 2003) cargoes HPV inhibition of nuclear transport of importin (Krawczyk et al. 2008; Nelson et al. viridae E5, L1 Importin-5 Papilloma cargoes 2003)

redistribution of lamin A/C causing nuclear Type 3 herniation, NPC clustered in herniations, 3 σ1s lamin A, NPC (Nup54 (Hoyt et al. 2004) Reo-virus might cause the observed G2-M cell cycle

Alteration of NPC/NE composition, modulation of Nuclear import/export of composition, NPC/NE of Nuclear modulation Alteration

Reoviridae block

e - growth defect in yeast probably through lack 4 HCV NS5A Importin-5 (Chung et al. 2000)

Flavi of nuclear import of ribosomal proteins

virida

- 51 -

Introduction

Viral Group Family Virus Targets Process Reference Component cleavage of Nups62/98/153 leading to Poliovirus Nup62, Nup98, Nup153 inhibition of importin-α/β and transportin-1 (Gustin & Sarnow 2001; Gustin 2003) mediated nucleo-cytoplasmic transport

other proteases?, NE Poliovirus 2A protease nuclear leakage (Belov et al. 2004) components? Rhinovirus Nup62, Nup153, Nup214, degradation leads to inhibition of nuclear (Ghildyal et al. 2009; Gustin & Sarnow

Picornaviridae 14 Nup358 import 2002) general inhibition of nuclear import and (Porter & Palmenberg 2009; Ricour et Cardiovirus L Nup62/98/153/358 export through hyperphosphorylation of al. 2009; Bardina et al. 2009) Nups

general inhibition of nucleocytoplasmic (Atasheva et al. 2010; Atasheva et al.

viridae VEEV H68 Importin-α/β, Crm1 transport by formation of a ternary complex - 2008) with importins and Crm1

Toga

(Alonso-Caplen et al. 1992; Fortes et al. CPSF, NXF1, p15, Rae1, E1B-AP5, - Influenza A NS1 inhibition of mRNA export 1994; Nemeroff et al. 1998; Chen et al. Nup98 1999; Satterly et al. 2007)

Ortho

myxoviridae Influenza A Nup62, NXF1, p15 increased mRNA and vRNA export (Morita et al. 2013) 5 (Reid et al. 2006; Reid et al. 2007; Filo- inhibition of STAT1 (Interferon signalling) Ebola VP24 Importin-α5/6/7 Mateo et al. 2010; Shabman et al. viridae and hnRNP C1/C2 nuclear translocation 2011) Rhabdo- Inhibition of specific import and export (Her 1997; von Kobbe et al. 2000; VSV M Nup98, Rae1 viridae (mRNA) pathways Enninga et al. 2002)

Rev Crm1 viral unspliced mRNA exported (Fornerod et al. 1997)

reduction and redistribution (NPCcytoplasm) of Nup62 leads to 6 HIV-1 Rev (as part (Monette et al. 2011; Monette et al. Nup62 inhibition of nuclear import of Transportin-1 of the vRNA) 2009) Retroviridae cargoes; Nup62 incorporation into nascent virions essential to vRNA export

- 52 -

Introduction

Viral Group Family Virus Targets Process Reference Component KTN1, NUP43, NUP58, NUP45, reduced levels of anchoring Nups Gp210 and

Nup50, NUP54, Nup62, GP210, POM121 alter NPC composition POM121 (only at NE not total), general composition changes of NPC; NUP107, NUP133, NUP155, redistribution of Nups (e.g. Nup153);

NUP160, NUP37, NUP85, NUP93 incorporation of some Nups into nascent (only at NE not total), Nup358 virions KPNB1, KPNA2, KPNA3, KPNA4, KPNA6, IPO5, IPO7, IPO9, TNPO1, (Monette et al. 2011; Chan et al. 2007; TNPO2, TNPO3, RANGAP1, RCC1, Reduction of these proteins that are Chan et al. 2009)

NUTF2, RANBP1, RANBP3, involved in nuclear import RAP1GDS1, SUMO1, RIP (only at NE not total) NUP35, NUP98, TPR, NUP153, NUP214, NUP93, POM121, RIP, increased expression to facillitate HIV

XPO1, XPO2, XPO5, XPO7, NXF-1 replication and likely export of vRNA (at NE)

- 53 -

Aims of PhD project

1.5 Aims of PhD project

Upon entry of HSV-1 capsids into the cell, they shed most of their tegument layer except for a few proteins which remain tightly associated during transport to the nucleus. Capsids of HSV-1 mutant tsB7 can be seen docking at the nuclear pore, a function which had been attributed to VP1-2. Further evidence showed that deletion of a NLS in the N-terminus of VP1-2 specifically prevented binding to nuclei. Although VP1-2 has diverged throughout evolution, some regions have been conserved even in the most distant relatives of HSV-1 suggesting that they might fulfil a similar function. This thesis is divided into four chapters which investigate the mechanism of HV nuclear docking using HSV-1 as a model:

(i) Investigation of putative NLS function in VP1-2 homologues from all HV sub-families (ii) Function of homologous NLSs and HSV-1 NLS mutantsin the context of HSV-1 replication (iii) Mechanism of nuclear transport of HSV-1 VP1-2 in the context of soluble proteins and incoming capsids (iv) Construction of a fluorescently tagged derivative of the VP1-2.ΔNLS virus to track incoming capsids and follow infected cells over time

Chapters one and two investigate the function of the VP1-2 NLS region in HSV-1 and that of a select set of homologues from human HVs. Briefly, I wanted to …

 define the best system to screen putative NLSs  narrow down the (NLS) region of interest using VP1-2 protein alignments and NLS prediction tools  find and characterise motifs from other HV sub-family members in nuclear transport assays using a heterologous protein (β-galactosidase) or truncated HSV-1 VP1-2  investigate the function of any functional motifs in HSV-1 replication by creating chimeric HSV-1 carrying functional NLS motifs from other sub-families  characterise the growth characteristics of any chimeric HSV-1  characterise the nature of the HSV-1 NLS motif in more detail  investigate the implications of mutations on nuclear localisation using a heterologous protein and HSV-1 VP1-2 nuclear localisation assay  characterise the impact of selected mutations on HSV-1 fitness

To test the hypothesis whether the N-terminus of VP1-2 of all HV sub-families comprises a functional NLS I cloned homologous VP1-2 regions of the representative HVs VZV (Varicella zoster virus, α), HCMV (β), EBV (Epstein-Barr virus, γ1) and HHV-8 (γ2) into the N-terminus of a heterologous

- 54 -

Aims of PhD project protein or into HSV-1 VP1-2. Their ability to confer nuclear localisation was tested in transfected cells to give a qualitative and quantitative measure of their function. To test the function of the homologous putative NLS motifs in replication they were transferred into the HSV-1 VP1-2 protein. This was achieved using a recombination-complementation assay based on the defect of the VP1-2.ΔNLS mutant. As such, any motifs that allow HSV-1 capsids to dock at the nuclear pore should be able to infect cells. Finally, I will correlate the data for nuclear transport for each homologous region with that of their ability to rescue HSV-1 virus replication. Recombinant (rescued) virus were isolated where possible and the recombinant viruses, which harbour homologous (NLS) regions of VP1-2, characterised in several cell lines. Chapter three investigates the mechanism of HV capsid nuclear docking using two biochemical approaches. I wanted to define the interaction partners of the VP1-2 NLS of HSV-1 and develop a system that, in theory, should enable identification of interacting partners of the VP1-2 NLS regions of any HV. Briefly, I will try to find VP1-2 interacting proteins with two pulldown approaches using…

 VP1-21-795 and its ΔNLS mutant  HSV-1 wild type and ΔNLS capsids

The first approach made use of a N-terminal fragment of VP1-2 as bait protein to identify the interacting proteins of the NLS. (residues 1-795). A system to select Stable cell lines inducibly expressing both the wild-type and the ΔNLS isoforms was developed which subsequently enabled co- precipitation of VP1-2 binding proteins. Interacting proteins were then analysed by tandem MS. This method was aimed to be a comparative analysis. Thus a mutant isoform lacking the NLS was included as a control and this should make a direct comparison of the interactome of wild-type and mutant isoform possible since nuclear transport factors should only bind to the wild-type protein. In an alternative second approach I attempted to optimise an in vitro method to mimic capsid entry, de-tegumentation and binding of cellular protein in a more ‘physiologic’ in vitro assay. This involved simultaneous de-envelopment using detergent, which would lead to the removal of tegument and binding of cellular proteins potentially relevant during entry. The binding proteins will ultimately be identified using MS. Comparing the list of proteins precipitated by either wild-type or ΔNLS capsids should give an indication of NLS binding partners in the context of capsids similarly to the first approach using soluble VP1-2. For chapter four I aimed to create a fluorescent derivative of the ΔNLS virus mutant comprising a fluorescently labelled tegument protein (VP16). Briefly, I wanted to …

 isolate GFP-VP16 recombinant VP1-2.ΔNLS viruses  characterise viral replication of the recombinant virus

- 55 -

Aims of PhD project

 study cellular behaviour using GFP-VP16 recombinant virus  find out if the defective VP1-2.ΔNLS virus spreads in rabbit skin cells

To achieve this, I purified recombinant viruses using a co-transfection approach. At first the growth characteristics were validated to confirm that the tags do not interfere with the virus life cycle. Subsequently, the GFP-VP16 recombinant virus was used to study the behaviour of infected cells over time and to find out why the defective ΔNLS virus appears to retain some rudimentary infectivity in rabbit skin cells.. To monitor virus spread in several cell lines, these were infected with complemented recombinant virus and images of the same area taken over several days. Additionally, to study cluster formation in rabbit cells, images of infected cells were taken every 20 minutes to elucidate how cluster formation was brought about.

- 56 -

Materials and methods

- 57 -

Materials and methods

2 Materials and methods 2.1 Materials

Table 4 Chemicals used for thesis. Most chemical compounds were obtained from Sigma or VWR unless otherwise stated.

Reagent Supplier 2x Taq polymerase master mix Sigma (P4600) 4′,6-Diamidin-2-phenylindol (DAPI) Sigma (D9564) Actinomycin D Sigma (A9415) agarose Invitrogen (18300-012) ammonium nickel(II) sulfate Sigma (574988) bromophenol blue Sigma (B-5525)

CaCl2 Sigma (C3881) cobalt(II) chloride hexahydrate Sigma (C8661) Collagen Type I (rat tail) First Link (60-30-807) Crystal violet (CV) Sigma (C0775) diaminobenzidine (DAB) Sigma (D8001) diazabicyclooctane (DABCO) Sigma (D2522) Doxorubicin Sigma (D1515) Doxycycline Sigma (D3447) DRAQ5 Life technologies (62254) DTT Sigma (GE17-1318-01) Dulbecco's Modified Eagle's Medium (DMEM) Gibco (41966-029) Dulbecco's Modified Eagle's Medium/Nutrient F-12 (DMEM F-12) Sigma (D6421) EDTA Sigma (E6758) EDTA-free protease inhibitor cocktail (tablets) Roche (11873580001) Foetal bovine serum (FBS) (Australia) Gibco (10099-133) geneticin Invitrogen (10131) glutathione covered sepharose beads Sigma (GE17-0756-01) glycerol Fisher (56-81-5) goat serum Simga (G6767)

H2O2 Sigma (216763) HEPES Sigma (H4034) Leptomycin B Sigma (L2913) L-Glutamine Gibco (25030-024) LiCOR block for Western blots LiCOR (927-40000) Marvel Milk Powder Tesco MEM non-essential amino acids (NEAA) Gibco (11140-035) MS grade trypsin Sigma (T6567) NaCl Sigma (S9888) N-bis(2-hydroxyethyl)-2-aminoethane sulfonic acid (BES) Merck (1152280025)

- 58 -

Materials and methods

Reagent Supplier New-born calf serum (NCS) (New Zealand) Gibco (16010-159) paraformaldehyde (PFA) Sigma (P6148-500G) Penicillin (100 U/ml) /Streptomycin (100 µg/ml) (Pen/Strep) Gibco (15140-122) Phenylmethanesulfonyl fluoride (PMSF) Simga (P7626) Phosphate buffered saline (PBS) (x10) Sigma (D1408-500ML) Phosphono acetic acid (PAA) Sigma (284270) piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES) Simgma (P1851) poly-L-Lysine Sigma (P2636) Polyvinyl alcohol (Mowiol 40-88) Sigma (324590) Pooled human serum (HS) Seralab (S-112-HS) Protein A sepharose Sigma (P9424) Protein G sepharose Sigma (P3296) Proteinase K Sigma (P2308) Q5 polymerase kit NEB (M0491S) Restriction enzymes NEB, Fermentas (various types) SDS Fluka (05030-1L) Silver stain kit Pierce (24600) sodium acetate Sigma (S2889) SYBR SAFE Invitrogen (S33102) Tetracycline approved FBS Gibco (631106) Tris-base Sigma (T1503) Triton X-100 Biorad (161-0407) Trypan Blue (0.4%) Invitrogen (T10282) Trypsin-EDTA (0.05%) Gibco (15400-054) Tween20 Biorad (170-6531) β-mercaptoethanol Sigma (M6250)

- 59 -

Materials and methods

Table 5 Antibodies used for this thesis.

Antigen Clone Supplier Species Concentration ICP8 1,10E+03 Abcam (20194) M 1:300 (IF) ICP4 10F1 Virusys Corp. (H1A021-1) M 1:500-800 (IF); 1:1600 (WB) VP16 LP1 T. Minson M 1:4000 (WB) gB R69 Gary Cohen R 1:10000 (WB) VP1-2 USP 1E12-E3 EMBL (MONTERONDO, ITALY) M 1:2000 (WB) VP26 P. Desai R 1:500 (WB) VP5 (ICP5) 3B6 Virusys #HA018 M 1:3200 (WB) UL37 C peptide P. Desai R 1:800-3000 (IF); 1:10000 (WB)

VP22 AGV30 Regal Group R 1:10000 (WB) UL25 V. Preston M 1:1000 (WB)

Primaries Invitrogen #R960-25 1:10000 (WB); 1:200 (IP); V5 epitope tag M 1:1000 (IF) GFP (better) Fitzgerald #20R-1769 R 1:10000 (WB) β-galactosidase Promega #Z3781 M 1:200 (IF) γ-tubulin Sigma #T5192 R 1:2000 (WB) ubiquitin FK-2 Enzo #BML-PW8810 M 1:2000 (WB) HA epitope tag 16B12 Covance #MMS-101R M 1:1000 (WB, IF) anti-DTX3L (BL3622) BL3622 Bethyl #A300-833A R 1:300 (IP) anti-DTX3L (BL3624) BL3624 Bethyl #A300-834A R 1:500-1000 (WB) 53BP1 CST #4937 M 1:100-200 Mouse-Alexa 594 Invitrogen (#A21203) D 1:1000 (IF) Mouse-Alexa 488 Invitrogen (#A11029) G 1:1000 (IF) Rabbit DyLight594 Thermo Scientific (#35560) G 1:1000 (IF)

Mouse Dylight 680 Pierce #35518 G 1:10000 (WB) Mouse Dylight 800 Pierce #SA5-10176 G 1:10000 (WB)

Secondaries Rabbit Dylight 680 Pierce #35568 G 1:10000 (WB) Rabbit Dylight 800 Pierce #SA5-10036 G 1:10000 (WB) 1:10000 (WB); 1:2000 Anti-Mouse-HRP Biorad G (Immunoperoxidase) 1:10000 (WB); 1:2000

Anti-Rabbit-HRP Biorad (Immunoperoxidase)

- 60 -

Materials and methods

2.2 Cell lines

COS (African Green Monkey fibroblasts) and Vero (African Green Monkey kidney cells) cells were maintained in Dulbecco’s modified eagle medium containing L-glutamine, pyruvate and 4.5 g/l D- glucose, supplemented with 10% neonatal calf serum and 50 μg/mL penicillin-streptomycin. HuH7 (human hepatocellular carcinoma cells), HEK293 (human embryonic kidney cells), RSC (rabbit skin cells) and HaCaT cells (immortalized human keratinocyte cell line) were maintained in DMEM supplemented with penicillin-streptomycin (50 μg/ml), 1x non-essential amino acid mix and 10% fetal bovine serum. For all HEK293 cells (including derivatives) flasks were coated with 10 μg/ml rat tail collagen type I and coverslips were coated with 200 μg/ml poly-L-lysine to allow for better attachment and retention during wash steps. RPE-1 (hTERT-immortalized retinal pigment epithelial cell line) were maintained in DMEM containing F12 nutrient mixture HAM supplemented with 10% FBS, 1x NEAAs, penicillin-streptomycin (50 μg/ml) and 5 mM glutamine. RSC-HAUL36 cells are a clonal cell line derived from RSC cells which express chromosomally integrated, N-terminally HA-tagged wild type VP1-2. Expression of the protein is maintained by supplementation of the media with 1 mg/ml geneticin. HS30 cells are a clonal cell line derived from Vero cells which express chromosomally integrated, N-terminally HA-tagged wild type VP1-2. Expression of the protein is maintained by supplementation of the media with 100 μg/ml geneticin. HEK293Tet-on cells (Clontech), a derivative of HEK293 cells constitutively expressing the tetracycline reverse transactivator (rtTA) protein, were maintained in HEK293 medium supplemented with tetracycline approved FBS and 100 μg/ml geneticin. The NT3 (VP1-2 N-terminal fragment 3) derivative cells pTH111 (NT3) and pTH112 (NT3ΔNLS) were additionally grown under puromycin selection (3μg/ml). Generally cells were split twice weekly at 1:5 to 1:10 ratios. For HEK293 cells (or derivatives thereof) flasks were coated with 20 μg/ml rat tail collagen type I in PBS for at least 30 min at 37 °C. If HEK293 cells (or derivatives) were plated for immunofluorescence (or any assay involving washing steps) coverslips or well were coated with 0.1 mg/ml poly-L-lysine for at least 30 min at 37 °C. Before plating cells all flasks or wells were washed once to remove excess collagen or poly-L-lysine.

2.3 Competent E. coli and transformation

To prepare competent XL1 blue cells (originally from NEB) a 1:100 diluted overnight culture was grown in LB for 2.5 h. Cells were pelleted at 2000 rpm for 10 min (4 °C) and resuspended in 100 mM ice cold CaCL₂. Cells were incubated for 45 min on ice, pelleted at 2000 rpm for 10 min at 4 °C and

- 61 -

Materials and methods resuspended in cold CaCl₂ supplemented with 16% glycerol. Cells were stored at -80 °C. For some applications commercially available low efficiency XL1blue or DH5α or high efficiency DH10 (e.g. for larger plasmids) E. coli were used. For transformation cells were defrosted on ice, mixed with the ligation mix (1 part ligation mix and 9 parts bacteria; see 2.9.1 for ligation protocol) and incubated for 30 min on ice. Cells were heat shocked at 42 °C for 45 sec and cooled on ice. Cells were recovered in lysogeny broth (LB) or SOC (super optimal broth with catabolite repression) medium under light shaking for 1 h before plating on LB plates containing the appropriate selection antibiotic. Transformation with 0.5-1 ng of plasmid in 100 μl of XL1 blue cells (home-made) gave rise to >100 colonies on LB plates containing appropriate antibiotics (>2x105 cfu (colony forming units)/g plasmid).

2.4 PCR

For most PCRs Q5 polymerase (NEB) was used. Generally the manufacturer’s recommendations were followed and the annealing temperature was determined using the NEB Tm calculator (see NEB website). To confirm certain types of constructs which could only be obtained with low efficiency and to enable screening of a large amounts of colonies, I used a pooled colony PCR approach. Briefly, transformed colonies were picked, streaked out on separate LB (+ antibiotic) plates and additionally grown in 300 μl LB (+antibiotic) in pools of 5-10 colonies per tube. These colony pools were incubated at 37 °C for 2 h and 1 μl was used instead of plasmid per 50 μl PCR reaction. To keep costs to a mimimum a 2x Taq polymerase master mix was used (Sigma). Annealing temperatures were estimated with the NEB Tm calculator. PCR products were separated on 1% agarose gels and detected using SYBR SAFE.

2.5 Plasmid preparation

Plasmid stocks were produced with Qiagen Mini- (Qiagen, #27106) or Maxi (#12163) prep kits and kept at 4 °C. 100 ng/μl stocks were prepared in double distilled (dd) H₂O which were used for transfections.

2.6 Transfection 2.6.1 Calcium phosphate protocol

This method of transfection was essentially performed as outlined in the original report (Chen & Okayama 1987) with minor modifications. Briefly, the final volume of the transfection mixture was made up in one tenth of the volume of the culture vessel (e.g. 250 μl for 2.5 ml using 6 well plates) and the cells were continuously kept at 5% CO2 (rather than the suggested 2-4%). Cells were seeded to

- 62 -

Materials and methods reach 50-70% confluency (3x105 RSC cells per well) by the next day (which was the time of transfection). For the rescue assay the medium was exchanged 2 h prior to transfection (2.5 ml in 6 well plates). The transfection mixture was made up freshly each time and left at room temperature for 20 min prior to addition to the cells. The mix was added drop wise and swirled gently. After 24 h the medium was exchanged to fresh 10% medium and left for another 48-72 h for cytopathic effect (CPE) to occur. If CPE occurred the cells were scraped into the medium and processed as outlined in the virus section. To create cell lines that inducibly express NT3 and NT3ΔNLS essentially the same transfection principles were applied with additional modifications to the reagents. The actual transduction was performed by Anna Stockum (laboratory of Dr Goedele Maertens) due to safety reasons. Briefly, a total of 20 μg of plasmid DNA (made up of 8 μg pCG-GagPol (Naldini et al. 1996),, 2 μg pVSV-G (Naldini et al. 1996), 10 μg pTH111 or pTH112) were made up with ddH2O to 450 μl. To this 50 μl 2.5 M CaCl2 were added and mixed well. This mix was subsequently dropped onto 500 μl 2x Hepes buffered saline (HBS)

(280 mM NaCl, 50 mM HEPES pH 7.1, 1.5 mM Na2HPO4) and left to equilibrate for 30 min at room temperature. The transfection mixture was added drop wise onto the cellular medium and the cells incubated with the DNA for 16 h. Cells were washed 2x with warm medium and the medium replaced with 5ml per 100mm dish. Virus was harvested at 24 h and 48 h post-infection (medium replaced at 24 h). The retrovirus was filtered through 0.45 μm cartridges and added directly to recipient HEK293TetON cells plated in 100 mm dishes at 1.5x106 cells/dish the day before. 48 h after the last transduction selection with 3μg/ml puromycin was started. Within one to two weeks single colonies grew out which were picked and expanded.

2.6.2 GeneJammer protocol

Cells were seeded in 12 well plates (containing glass coverslips if samples were to be used for immunofluorescence studies) at 10⁵ cells/well in 1 ml media. Transfection was undertaken with COS or HuH7 cells using the manufacturer’s instructions of the Gene Jammer kit (Agilent; 204131). 3:1 ratio of Gene Jammer transfection reagent to DNA was used as standard with a total of 500 ng per 100 μl transfection mixture being applied per well. Cells were fixed 24 h post-transfection (unless otherwise indicated) with 3.7% paraformaldehyde (PFA)/PBS or Methanol or harvested in 1x SDS sample buffer (50 mM Tris-HCl, 2% SDS, 3% glycerol, 5% β-mercaptoethanol, 0.1% bromophenol blue) for Western blot.

- 63 -

Materials and methods

2.7 SDS PAGE, Western blot, Coomassie and silver stain procedure

Cells were washed in ice cold PBS and harvested directly from 6-24 well plates in Laemmli buffer (0.05 M Tris base, pH 7.0, 2% SDS, 5% β‐mercaptoethonol, 3% Glycerol and 0.005% Bromophenol Blue). If larger plates were used then cells were scraped into 5-10 ml PBS and sedimented before lysis. If the sample was very thick due to DNA release they were sonicated briefly (up to 30 sec in a water bath sonicator). The samples were heated to 95 °C for 3-5 min and stored at -20 °C or used straight away. Dependent on the molecular weight of the proteins of interest they were separated using standard running buffer (25 mM Tris base pH 8.3, 192 mM Glycine, 0.1% SDS) at 150 V on 12%, 10%, 8%, 7% PAGE gels or 3-8% pre-cast gradient SDS-polyacrylamide gel electrophoresis (PAGE) mini gels (>150 kDa) (NuPAGE; # EA03752BOX) using the SureLock system from Invitrogen. For 3-8% pre-cast gels a Tris-MOPS buffer was used (Invitrogen, 50 mM MOPS, 50 mM Tris base, 0.1% SDS, 1 mM EDTA, pH 7.7). Gels were either stained with coomassie solution for 30 min (0.1% Coomassie Blue, 10% Acetic acid, 50% Methanol and 40% H2O) followed by overnight de-staining (10% Acetic acid, 40% Methanol and 50% H2O), silver stained (see manufacturer’s instructions for silver stain details; Pierce #24600) or wet-blotted at 250 mA for 2 h or 72 mA for 16 h onto nitrocellulose membranes using transfer buffer (25 mM Tris base, 192 mM Glycine, 10% Methanol). Membranes were stained with Ponceau S solution, blocked with 0.5x LiCOR blocking solution (as 1:1 mix with PBS), probed in the dark with primary and secondary antibodies diluted in PBS supplemented with 0.1% Tween20 (PBST) and 5% skimmed milk or 1.5% BSA. After each probing step membranes were washed 3x 15 min in PBST and once in PBS. Membranes were dried and read with the LiCOR Odyssey imager and stored at room temperature in the dark. Image analysis and preparation was done with Image Studio Lite software. Generally images were exported as tagged image files (tif).

2.8 Immunofluorescence microscopy

Cells were treated as described. For immunofluorescence analysis samples were usually grown in 12 well plates containing glass coverslips (+/- poly-L-lysine). Cells were washed once in ice cold PBS and fixed in 4% PFA for 20 min (at room temperature) or 100% methanol (at -20 °C) for 5-10 min. Cells were washed 3x in PBS. If PFA was used cells were subsequently blocked and permeabilised for 1 h with PBS/0.1% Triton X-100 (PBSTx) containing 300 mM glycine and 10% NCS, FBS or Goat Serum. Samples were probed with primary and secondary antibodies in a dark, humidified chamber using block buffer (no detergent). DAPI (4′,6-Diamidin-2-phenylindol, 0.5-1 μg/ml) or 1, 5-bis{[2-(di- methylamino)ethyl]amino}-4, 8-dihydroxyanthracene-9, 10-dione (DRAQ5, 0.5-1 μM) were used as DNA counterstains as part of the secondary antibody mixture. Coverslips were washed 3x in PBS and

- 64 -

Materials and methods ddH₂O, dried (70% Ethanol) and mounted in Mowiol (10% Mowiol 4-88, 25% glycerol, 100 mM Tris- HCl) containing 2.5% DABCO (1,4-diazobicyclo-[2,2,2]octane). Images were taken with Image Pro Plus software on an Axiovert M135 inverted microscope or the LSM 500 Pascal or LSM510 Meta confocal laser scanning microscopy setup using 488nm, 543nm and 633nm lasers.

2.9 Generation of plasmids 2.9.1 NLS β-galactosidase constructs

NLS sequences were synthesized as oligo nucleotides (Sigma; see appendix) to generate in- frame fusion proteins with β-galactosidase (β-gal) as outlined elsewhere (Abaitua & O’Hare 2008; Hennig et al. 2014). Briefly, the β-gal vector contains two restriction sites, HindIII and BglII, immediately adjacent to the Kozak and translation initiation sites. Two complementary oligos per NLS were denatured, annealed and digested as follows: Oligo nucleotides were diluted in ddH₂O to a concentration of 10 or 100 μM followed by denaturation at 96 °C for 10-15 min and cooling on ice.

Annealing mixtures were prepared as 50 μl reactions containing 1x NEBII buffer, ddH2O and 3 μM of each oligo nucleotide. To anneal the reactions were heated to 98°C (from ice) and the temperature dropped by 1 °C per min until 25 °C was reached. Reactions were digested with 20 U HindIII and BglII enzymes for 5 h and purified using a PCR reaction clean-up kit (Qiagen; # 28104). NLSs were ligated at 14 °C for 4 h with 6:1 excess over vector into 100 ng HindIII/BglII digested β-gal plasmid (6 x [insert in bp]/[vector in bp] x 100 ng). Plasmid backbone was digested with 10 U each of HindIII and BglII in 20-30 μl reactions using NEBII buffer. 1 μl calf intestinal phosphatase (CIP) were added during the last hour of the digestion. The backbone was isolated from 1% agarose gels under blue light (SYBR safe) by excision and DNA was purified using a QiaEXII kit (Qiagen) according to manufacturer’s instructions. This same procedure was used for all other gel purifications. R1 constructs were created by chemical synthesis (GeneArt). Each R1 region contained from its 5’ end in the following order (without DNA spacers): HindIII site, Kozak sequence, ATG codon, HA tag, XhoI site, R1 region, NheI site, BglII site. For insertion into the β-gal vector the HindIII/BglII restriction strategy was used (see above). The sequences were obtained from NCBI (see comparison of in silico NLS prediction algorithms for accession numbers). All constructs were sequenced using Beckman and Coulter Sanger sequencing using specific primers in the N-terminus of β-gal (see Appendix 1: List of primers with descriptions).

- 65 -

Materials and methods

2.9.2 Creation of NT6 constructs

The cloning strategy is outlined elsewhere (Hennig et al. 2014). Briefly, chemically synthesised R1/R8 regions flanked by XhoI and NheI sites were inserted into the intermediate vector pTH19 between these two sites. This vector contained the engineered BamHI-MluI fragment of NT6v2 with unique XhoI and NheI sites (silent mutations were introduced leading to no primary amino acid sequence changes). BamHI-MluI fragments were subsequently inserted into NT6v2 (between BamHI/MluI) and sequenced using Beckman and Coulter sequencing services (see appendix for primers).

2.9.3 Creation of mammalian GST constructs

An existing GST-NT4 construct, harbouring an EcoRI-EcoRI fragment from the N-terminus of VP1- 2 (around 140 amino acids around the NLS), was used as the basis of my cloning procedure. GST from pGEX-6P was cloned into pCAGGS (see Addgene for a map) to allow for mammalian expression and the NT4 part was cloned into this by EcoRI-EcoRI ligation (in frame). To create NLS bearing GST fusion proteins, the NT4 part was removed and the pCAG-GST backbone used for insertion of digested PCR products. The PCR templates were pTH19 derived intermediate vectors which contained the chemically synthesised R1s (HSV-1 wild-type, K428A and ΔNLS) with flanking HSV-1 sequence (i.e. as part of the BamHI-MluI fragment). These will still be referred to as R1 regions in the text. The ΔNLS mutation had to be transferred from NT6.ΔNLS into the pTH19 backbone (direct BamHI-MluI insertion). The PCR product was created using the primer pair prTH101/102 and contained a 5’ MfeI and a 3’ EcoRI site which were used for the cloning procedure (insertion into pCAG-GST). Constructs were sequenced using Beckman and Coulter’s universal GST primers.

2.10 Nuclear import assay (β-galactosidase, GFP and VP1-2 constructs)

The procedure was the same as outlined in a previous report (Hennig et al. 2014). Briefly, HuH7, HEK293 (with poly-L-lysine) or COS cells were seeded on 13-16mm coverslips at 1.5x105 cells/well 24 h prior to transfection. The next day cells were transfected (see 2.4) with 500-1000 ng of a plasmid in 12 well plates. Cells were fixed 24 h post-transfection with 4% PFA and immunostained (see 2.5) using mouse anti-β-gal or anti-V5 antibody and secondary goat anti-mouse 488 labeled antibody. DNA was counterstained with DAPI or DRAQ5. Coverslips were mounted in Mowiol contained 2.5% DABCO. Images were taken at 10x, 20x, 40x or 63x magnification using fluorescence or confocal microscopes. For the nuclear export inhibition assay cells were transfected with NT6 constructs and with the additional control constructs GFP, GFP-NLS (CMV.R1) and GFP-NLS-NES (CMV.R1 and Rev NES) for 24 h.

- 66 -

Materials and methods

Cells were then treated with 10ng/ml leptomycin B for 3h prior to fixation. Controls were treated with vehicle only (0.1% DMSO). For inducible cell lines (NT3), cells were seeded at 1.5x105 cells/well on poly-L-lysine (Sigma) covered 13-16mm coverslips and treated with 1μg doxycycline for a minimum of 19h (up to 36h).

2.11 Creation of inducible NT3 and NT3ΔNLS cell lines

The series of vectors created for this part were based on the retroviral backbone (Figure 6) created by Dr. Goedele Maertens (Maertens et al. 2010). Initially, the BglII to EcoRI part was excised which included the CMV promoter and the multi cloning site (MCS). This was replaced with the PCR product of prTH89/90c (BamHI-EcoRI). The PCR product contained the pTRE-tight cassette (pTight vector from Clontech) and a designed MCS flanked by 5’ BamHI and 3’ EcoRI sites. The 5’ BglII-BamHI ligation lead to loss of either site upstream of the pTight promoter. NT3 (and ΔNLS) were PCR amplified using the primer pair prTH93/94 and the purified product was digested with NheI and EcoRI. The resulting restriction fragment was inserted between XbaI and EcoRI sites of the new MCS. Vectors were tested by transient transfection for inducibility with doxycycline and Western blot. Cell lines were created as outlined in 2.5.1. Since the puromycin cassette was preceded by an IRES the selection was less efficient since leaky expression was required. Nonetheless, at least one colony of transduced cells was obtained for either construct. Cells were tested with Western blot against V5 (tag at the N- terminus of NT3).

- 67 -

Materials and methods

Ssp I (6856) CMV/MSV

Sca I (6532) Asc I (543) AmpR R U5 Psi

Bgl II (1577)

Alw NI (5575) Xba I (1596) pQFlag CMV promotor 7200 bp Not I (2241)

Xcm I (2259)

SV40 ori Mfe I (2260) SV40 prom Flag 3' MoMuLV LTR (deletion in U3) Age I (2290)

Nhe I (3937) Pin AI (2290)

Pvu II (3838) Xho I (2308)

Eco RV (3795) Bam HI (2314)

Bsp EI (2317)

Eco RI (2321)

Apa I (2446) IRES

Pml I (2645)

Bsp MI (2668)

PflMI (2782)

Rsr II (3050) Puro R

Figure 6 Retrovirus vector for creation of pTH111 (NT3) and pTH112 (NT3ΔNLS). (Maertens et al. 2010)

2.12 Immunoprecipitations

To perform most immunoprecipitations (IPs) a PBS based buffer system was used. PBS (pH 7.4) was supplemented with 1% Triton X-100, 1 mM EDTA, protease inhibitor cocktail (1 tablet per 10 ml), 10 mM DTT and 1 μM Phenylmethanesulfonyl fluoride (PMSF). Any modifications (such as during capsid preparations) are mentioned in the text. The method was adapted for the size of the vessels used (e.g. 6 well plate for NT6 IP, 100 mm dish for NT3, 550 cm2 dish for GST-R1s) For NT3 IPs cells were seeded on poly-L-lysine coated 100 mm dishes at 1.5x105 cells/dish 16-24 h prior to induction. Cells were induced with 1μg for a minimum of 19 h but for a maximum of 36 h. Cells were washed twice in ice cold PBS and lysed in 1 ml of ice cold lysis buffer for 30 min at 4 °C under gentle agitation (all steps from now on at 4 °C or on ice). A small amount of this total sample was diluted with Laemmli buffer (Total). Debris were sedimented at 13000 rpm for 10 min. The lysate was pre cleared with 25 μl protein A bead equivalent volume for 1 h under agitation, beads sedimented and the Input transferred to a new tube. A small sample was diluted with Laemmli buffer (Input). NT3 was bound with anti-V5

- 68 -

Materials and methods antibody (1:200) for 2 h under agitation before incubating with 25 μl protein A beads for 1h. Beads were sedimented at 6000 rpm for 15 sec and a small sample of the unbound supernatant diluted in Laemmli buffer (unbound). Beads were washed 5x with lysis buffer. Proteins were either eluted with Laemmli buffer or subjected to on-bead digestion with trypsin.

2.13 On bead digestion of IPs for mass spectrometric analysis

NT3 and NT3ΔNLS expressing cells or HEK293 cells were plated in coated 100 mm or 530 cm2 dishes, respectively. Cells were either induced with 1 μg/ml doxycycline or transfected with 90 μg GST- R1, GST-R1.K428A or GST-R1.ΔNLS. Cells were lysed and IPs performed as described. GST fusion proteins were immobilised with glutathione covered sepharose beads. After the final wash of the IP procedure, the buffer was aspirated and the beads resuspended in one volume (use bead equivalent volume to calculate) of 50 mM ammonium bicarbonate buffer (pH 8-8.5). Approximately 0.5 μg MS grade trypsin was added per 20 μg of substrate. The digestions were incubated for 16 h at 37 °C under vigorous shaking to keep beads resuspended. Peptides were analysed by tandam MS in collaboration with Dr Remi Serwa (South Kensington Campus).

2.14 In silico protein analysis 2.14.1 Protein alignments, NLS, NES and Ubiquitination prediction

To find regions of conservation protein sequences were retrieved from Uniprot and inserted in FASTA format into text files. These were then imported into Clustal Omega and aligned. Sequences were then manipulated manually with GeneDoc software and exported as bitmap files. To predict putative NLSs in the N-terminus of VP1-2 either just the R1 region or the first 1000 amino acids were used. The following five online tools were used to make sure putative NLS sequenced are not missed: NLSmapper, NLStradamus, Euk-mPLOC 2.0, NucPredict and CELLO (Yu et al. 2006; Nguyen Ba et al. 2009; Chou & Shen 2010; Kosugi et al. 2008; Kosugi, Hasebe, Matsumura, et al. 2009; Kosugi, Hasebe, Tomita, et al. 2009; Brameier et al. 2007). For this amino acid sequences were uploaded in FASTA format and the results collated in a table (see results section). Nuclear export signal (NES) and ubiquitination prediction was done in similar fashion using the online tools NetNES1.1 (La Cour et al. 2004) and UbPred (Radivojac et al. 2010), respectively.

2.14.2 Analysis of MS data

Data was received as a spreadsheet and the lists of proteins were manually filtered using Microsoft Office Excel functions. Protein hits were included if they satisfied the following criteria: (i) at

- 69 -

Materials and methods least two unique peptide are identified (ii) peptides at least four fold enriched in either sample compared to the negative control (IP from the parental HEK293TetON cell line).

[Absolute protein abundance (log2) in sample – Absolute protein abundance in control] >= 2

This way 199 proteins were identified which were subdivided into different categories as outlined in the text. Annotations (localisation and main functions) were retrieved from Uniprot (or where necessary from GeneCards) using the protein IDs supplied in the MS data. For example, for DTX3L (ID: Q8TDB6) ‘DNA repair’ was the retrieved function while ‘ubiquitin-protein transferase activity’ was not regarded as a function for the purpose of this analysis. Some proteins were allocated multiple functions which is represented in the analysis (i.e. more annotations than proteins). Although this introduced subjective bias it was required to simplify the analysis and representation. For graphical illustration Prism 5 software was used. To obtain enrichment scores the freely accessible DAVID tool was used (Huang, Brad T Sherman, et al. 2009; Huang, Brad T. Sherman, et al. 2009). Gene lists were uploaded and Fisher exact scores used for the purpose of illustration of significant enrichment. Since many different annotation terms are used almost interchangeably but not consistently (e.g. nuclear, nucleoplasm, nucleus, nuclear part, etc.), I combined the gene lists for each term that had a Fisher score below 0.00001 (i.e. highly enriched).

2.15 DNA damage assay

HEK293TetON cells inducibly expressing NT3 or NT3ΔNLS (and parental control) or U2OS cells were seeded onto poly-L-lysine coated coverslips to reach 50% density at the time of induction with 0.1-1 μg/ml doxycycline or transfection with NT3 plasmid DNA. Cells were induced or incubated with plasmid DNA for 24 h prior to addition of 100 nM doxorubicin (or 0.1% DMSO) to induce DNA damage. Cells were treated for either 4 or 24 h before fixing in 4% PFA (U2OS cells only for 24h) as described before. Cells were stained with anti-53BP1 antibody as described in section 2.8 and imaged with an inverted fluorescence microscope. For quantification images were taken at 10x magnification with the same settings for each condition. Contrast and brightness were enhanced using the Fiji plugin of Image J. Individual 53BP1 foci were reduced to single black pixels using the ‘find maxima’ function of Fiji and DAPI stained nuclei (as ROIs) were transferred to the resulting binary 53BP1 images. Dots per nucleus were counted manually. Since without DNA damaging agents the number of foci was generally lower than 10 per nucleus this was taken as a threshold for induced DNA damage.

- 70 -

Materials and methods

2.16 Protocol for virus culture and infections 2.16.1 Virus infection

Cells were seeded in 6, 12, 24 or 48 well plates at 3x10⁵, 1.5x10⁵, 8x10⁴ or 3x10⁴ cells/well one day prior to infection, respectively (three times more for HaCaT cells). Media was removed and cells incubated with inoculum for 1-2 h on a plate shaker at 37 °C. Inoculum was aspirated and 2% FBS media added (unless otherwise indicated).

2.16.2 Plaque assays

Cells were seeded at 3x10⁵ per well one day before. Virus titres were determined with 10 fold serial dilutions. The standard concentration range was 10-3 to 10-8. Infection was done as described above. Media was changed to DMEM/2%FBS supplemented with 2% human serum. Cells were fixed with 4% formalin and stained with crystal violet (CV) 3d post-infection and plaques counted at 6.3x magnification. For plaque size measurements, especially of the smaller recombinant virus plaques, the staining protocol was modified from a previous report (Green et al. 1989). Infected cells were fixed as described for immunofluorescence. Essentially the staining protocol was the same as used for conventional immunofluorescence microscopy but it was performed directly in 6 well plates (rather than on coverslips) and samples were not mounted. Primary antibodies of choice were polyclonal rabbit anti- UL37 (1:3000) or monoclonal mouse anti-ICP4 (1:800) used in combination with horseradish peroxidase (HRP) conjugated goat anti-rabbit/mouse secondary antibody (1:2000). After the last wash the substrate solution (0.05% diaminobenzidine [DAB], 0.05% cobalt Chloride, 0.05% nickel ammonium sulfate and 0.015% hydrogen peroxide in PBS, pH 7.2) was prepared fresh from 20x concentrated chemical stocks and 10x concentrated PBS. The sample well were aspirated and 1 ml substrate per well was added and incubated until plaques were visible (up to 10 min). Plates were washed in PBS and could be dried for storage. Images of individual fields were obtained with Q capture Pro 6.0 software using a Leica Wild M420 binocular equipped with a makrozoom wild objective (#376777) an Imaging Go-3 camera at 6.3x magnification. Alternatively, tiled images were acquired at 10x magnification (Image Pro Plus software, AFA module). Plaque size was measure using Quantity One 4 software using the image contour tool.

2.16.3 Single step growth curve and protein expression analysis

For single step growth curves, cells were seeded in 12 or 24 well plates to reach 100% confluency at the time of infection (next day). Cells were infected at MOI of 5 for 1 h, acid washed (citric acid, pH 3.0) for 1 min and incubated with 2% FBS containing medium for the desired time points. Cells and

- 71 -

Materials and methods medium were harvested (as total virus produced) by freeze thawing and titrated on both RSC and RSCUL36 cells. For protein expression assays, the infected cells were scraped into the medium and sedimented at 3000 rpm for 5 min. The cell pellet was washed once in PBS and the cells lysed in Laemmli buffer.

2.16.4 Multi step growth curve

For multi-step growth curves, cells were seeded in 12 or 24 well plates and infected at 100% confluence with a MOI of 0.001 as above. The infection was left to progress in 2% FBS containing medium. Monolayers were harvested at 0, 24, 48 and 72 h post-infection (hpi) as above and titrated on both RSC and RSCUL36 cells.

2.16.5 Virus stock production

Virus stocks for KOS wild type, KOS Bacmid, K.ΔNLS, K.ΔNLS.R, K.ΔNLS.GFP-VP16, strain 17 wild type, strain 17.VP16-GFP and strain 17.GFP-VP16 were created in RSC, Vero, RSCUL36 or HS30 cells (all ΔNLS versions). Working stocks were made by infecting 2x10⁸ cells at a multiplicity of infection (MOI) of 0.001-0.005 (plaque forming units/cell or PFU/cell). Media and cells were harvested 3-4d post- infection at 100% cytopathic effect, cells sedimented at 3000 rpm at 4°C for 30min and the media/supernatant spun at 19000 rpm at 4°C for 90min. The pellet of this media/supernatant was resuspended in RSC media (1% FBS) at 4°C overnight and used as extracellular virus stock. The cell pellet of was resuspended in 10 ml media (1% FBS) and cell-associated virus released by three freeze- thaw cycles. Debris were sedimented by spinning at 3000 rpm for 15min at 4°C and the supernatant stored as cell associated virus stock. Each stock was titrated on the appropriate cell line (RSC or RSCUL36).

2.16.6 Viral genomic DNA production

2x10⁸ RSC or RSCUL36cells were infected at MOI of 0.001-0.005 and only the medium was harvested at 100% cytopathic effect. The edium was spun at 3000 rpm at 4°C for 30’ to sediment cells/debris. The supernatant was spun at 19000 rpm at 4°C for 90min to sediment virus particles and the pellet resuspended in buffer (10 mM Tris HCl pH 7.6, 1 mM EDTA, ddH₂O to make 4.75 ml) overnight. The pellet was resuspended the next day and mixed with SDS (1% final) and proteinase K (50 μg/ml. The mixture was incubated at 50°C for 4h and the DNA was subsequently phenol-chloroform purified (Sambrook and Russell, 2006). Purified DNA was precipitated overnight at -20°C with 500 μl

- 72 -

Materials and methods

3M Sodium acetate and 12.5 ml 100% ethanol. DNA was sedimented at 3500 rpm at 4°C for 30min. The pellet was washed with 18.75 ml 70% ethanol, air dried and resuspended in 200 μl sterile ddH₂O.

2.16.7 K.VP1-2.ΔNLS virus rescue assay

For the rescue assay, RSC cells were seeded at 3x10⁵ in 6 well plates one day prior to transfection. 1 μg K.VP1-2.ΔNLS DNA and 1 μg of the different NT6 plasmids were transfected using the calcium phosphate procedure (Chen & Okayama 1987). Cells were fixed or harvested 3-4d post- transfection. Fixed plates were stained with crystal violet (0.1% CV, 10% methanol in ddH2O) and plaques counted and compared to the positive control (HSV-1 VP1-2 wild type). For yield comparisons, instead of CV staining plates, the monolayers were scraped into the medium and everything collected. The mixture was lysed by freeze-thawing 3x (switching between dry ice ethanol bath and 37°C). Debris were sedimented at 3000rpm for 10’ and the supernatant titrated on RSCUL36 cells. To isolate viruses, the lysate was titrated on non-complementing RSC cells. The procedure was essentially the same as for VP16-GFP recombinant virus (see below) just that more plaques were picked since there was no marker. Since the background of this assay was extremely low, this was deemed sufficient to obtain recombinant viruses. Recombinant viruses were confirmed by amplifying the region around the BamHI-MluI inserts with PCR (Q5 proof reading polymerase) and sequencing the PCR products (Beckman and Coulter).

2.16.8 Generation of VP16-GFP and GFP-VP26 recombinant viruses

Recombinant viruses were created using the Calcium phosphate transfection protocol of viral genomic DNA and plasmid DNA (containing appropriate, homologous flanking regions). RSC or RSCUL36 (for ΔNLS) cells were seeded the day before at 5x10⁵ cells/60 mm dish. Media was changed 2h prior to transfection to 2.5 ml media. A 250 μl transfection mixture contained the following mixed in order: 1x BES buffer (50 mM BES [N-bis(2-hydroxyethyl)-2-aminoethane sulfonic acid;, 140 mM NaCl, 0.75 mM Na₂HPO₄), 1 μg viral DNA, 4 μg plasmid, H₂O (to make 250 μl) and 125 mM CaCl₂. Media was changed to DMEM/2% FBS 24 h post-transfection. Cells and media were harvested at 100% cytopathic effect, freeze-thawed three times in ethanol/dry ice and titred on complementing or non- complementing cells (see 2.8.5). For purification of recombinant viruses each titration was overlaid with 0.4-0.6% agarose (included in DMEM/2%FBS) and green plaques picked 3 days post-infection (dpi) with a blunt ended needle. The titration procedure was repeated until a sufficiently pure stock was obtained.

- 73 -

Materials and methods

Figure 7 Flow chart to create and isolate a recombinant HSV-1 virus. (A) Viral genomic DNA (K.ΔNLS or revertant K.ΔNLS.R) was co-transfected with GFP-VP16 encoding plasmid DNA. (B) Homologous recombination co-transfected cells. The GFP-VP16 plasmid encodes the appropriate flanking regions. (C) The lysate from one co-transfection is used as inoculum to infect a monolayer of complementing cells overlayed with 0.4-0.6% agarose containing medium. In sequential steps of plaque picking and re-infection cycles the population can be isolated and enriched on the basis of its green colour.

2.16.9 Time lapse microscopy of K.VP1-2.ΔNLS.VP16-GFP infected cells

RSC cells were seeded in 6 well plates and infected the next day (100% confluency) with approximately 150 pfu/well as described before (with minor modifications). The medium used was

DMEM F-12 (15 mM HEPES; see RPE-1 medium) instead of regular DMEM since it does not require CO2 for buffering. After the infection the cells were transferred to a temperate mini-chamber which sat directly on the stage of an inverted fluorescence microscope. Individual fields of infected cells were picked which would be imaged over the course of 48 h. Images were acquired every 20 min for 48 h (in two parts) using Image Pro Plus software (AFA module). Each image corresponds to one frame of each avi movie file.

- 74 -

Materials and methods

2.16.10 Capsid extraction 2.16.10.1 Preparation of clean viral stocks

KOS bacmid strain virions were produced (after optimization) from RPE-1 (high yield of intracellular virus at 3 dpi) or HaCaT (slightly higher yield for extracellular virus at 3 dpi) cells (RSC and Vero cells were also tested but produced less virus). Briefly, cells were seeded in 800 cm2 roller bottles in HEPES buffered medium (DMEM F-12 supplemented with PS, NEAAs, 4mM L-glutamine and 10% NCS [RPE-1] or 10% FBS [HaCaT]) due to lack of CO2 ventilation. At 90-100% confluency (around 2- 4x108 cells/bottle) cells were infected at a MOI of 1 and the cells harvested by scraping into the medium 24 h later (100% CPE). Cells and debris were sedimented at 300x g and the supernatant separated from the cellular pellet. Supernatant associated virus (SN) was collected by high speed centrifugation at 19000 rpm for 1 h at 4 °C and subsequent ultracentrifugation of the resuspended pellet through a 35% sucrose (in PBS) cushion. The pellet was resuspended in 200-500 l PBS or water. Cellular virus was liberated from cells on ice by incubation with TNE (Tris pH 8, 500 mM NaCl, 1 mM EDTA) for 30 min under periodical agitation. The resulting procedure was repeated once. The cellular debris were sedimented at 3000 rpm for 10 min and supernatants pooled and further sedimented through 35% sucrose (= TNE stock). For test purposes the left-over cell debris were freeze thawed 3x and sedimented through 35% sucrose. Virus was titrated on RSC cells.

2.16.10.2 Capsid extraction

Purified virions were extracted using different buffer systems (see 5.2.2) but analogous to the following method: 50 l virions in PBS (for some applications virions were produced in water for buffer compatibility) were mixed with 150 l of PBS (or other) and 200 l of 2x concentrated extraction buffer (here PBS containing 1% TX-100, 2 mM EDTA, 2x protease inhibitors, 2 μM PMSF). For some applications, 200l of HEK293 cell extract of 2-4x106 cells was added as 2x extraction buffer. For this 107 HEK293 were washed in PBS, scraped and extracted using 500 l PBS (or piperazine-N,N′-bis(2- ethanesulfonic acid) (PIPES)) buffer containing 1% TX-100, 2x protease inhibitors and 2 μM PMSF for 30 min. Lysates were cleared for 30 min at high speed (25000 rpm) using a pre-cooled table top ultracentrifuge (rotor TLA-45) and low binding Eppendorf tubes (Sigma, # Z666505). The extraction reaction was incubated on ice for 15-60 min without significant differences in degradation. Finally, the soluble fraction was divided from the capsid containing fraction by sedimentation (with or without a 35% cushion) for 30 min (or 45 min if through a cushion) at 25000 rpm. Soluble proteins (S fration) were aspirated and the pellets (P fraction) washed once in extraction buffer. The final pellets were resuspended in 1x Laemmli buffer and analysed using SDS-PAGE and silver staining or Western blotting. A small amount of supernatant was kept and diluted with 4x Laemmli buffer (soluble fraction).

- 75 -

Results

- 76 -

Identifying NLSs of VP1-2 homologues

3 Identifying NLSs of VP1-2 homologues 3.1 Background

VP1-2 is an essential tegument component which is conserved across all HVs (Kattenhorn et al. 2005; Kelly et al. 2009). Although it was shown to be required during assembly (Desai 2000), transport and nuclear docking of nucleocapsids (Abaitua et al. 2012; Roberts et al. 2009), parts of the protein appear to be dispensable for growth in cell culture without deleterious effects on growth characteristics (Böttcher et al. 2008; Bolstad et al. 2011; Böttcher et al. 2006). Several regions were shown to facilitate the essential functions of VP1-2. However, which part was required for docking at the nucleus, was, until recently, not known (Abaitua et al. 2012). For regular proteins to reach nuclear pores, they require NLSs and our laboratory has previously shown that a stretch of four basic residues was absolutely required for VP1-2 (HSV-1) nuclear localisation (Abaitua & O’Hare 2008) and virus docking at the nuclear pore while being dispensable for nucleocapsid assembly (Abaitua et al. 2012). This patch was also sufficient to confer nuclear localisation to a large heterologous protein (Abaitua & O’Hare 2008). It appeared reasonable to assume that the NLS in the VP1- 2 N-terminus recruits a nuclear transport factor directly to fulfil its function during entry and thus identifying the regions with the same function in other HVs could aid elucidation of how these might traffic to the NPC. Identification of NLS types in other HVs could lead to elucidation of their mode of nuclear docking which, theoretically, could lead to the identification of drugable entry pathways that may be used to lower the health risks posed by HVs in immunocompromised patients such as transplant recipients or AIDS patients. My hypothesis was that VP1-2 homologues from all HV sub-families contain a NLS at an analogous position, that is, the first basic patch adjacent to the N-terminal USP domain. To empirically determine if there could be functional NLSs in the primary sequence other VP1-2s a two-step analysis of VP1-2 sequences was performed. Initially, the representative protein sequences of the VP1-2 homologoes of all HV sub-families were aligned based on the conservation in the USP domain and the body of VP1-2. Since the HSV-1 NLS was found in the N-terminus this narrows down the initial search to a similar region of the homologues. Subsequently, to identify putative NLSs in the aligned sequences several NLS prediction algorithms should enable me to decide which specific areas to test in nuclear transport assays.

3.1.1 Sequence alignment of the VP1-2 NLS regions from all HV sub- families

The first step to identifying putative NLS regions was to align areas within the primary sequence of VP1-2 homologues that show similarity. Due to the high degree of conservation between VP1-2 proteins, I assumed that each linker between USP and middle segment, despite poor overall homology, contains

- 77 -

Identifying NLSs of VP1-2 homologues similar activity. Both the USP and middle segment are relatively easy to spot using available protein alignment tools. The residues linking these two domains are only poorly conserved overall. The most N- terminal basic regions of this linker were aligned (Figure 8). The alignment highlights a common basic character (red background, white letters) and subgroup-specific conservation (see Figure 8, A through D). In all linker segments additional enrichment of prolines (grey shading) can be identified. This indicates that this region may not form intricate secondary structure due to the negative effect of prolines on both helices and beta sheets (Chou & Fasman 1974). Additionally, all equivalent regions feature an abundance of phosphorylatable residues (serines, threonines) and acidic residues, which could form so called PEST sequences. Such regions intrinsically destabilise secondary structure (Dunker et al. 2001) and thus would aid the presentation of any NLS in that region as linear epitopes. Four distinct, mostly sub-family specific patterns can be grossly distinguished. (i) Within the -HVs (Figure 8A) two basic regions, designated R4 and R5, are divided by a long, well conserved spacer enriched in potential phosphorylation sites (serines and threonines; yellow background). Despite the significant conservation at the N-terminus of the spacer the spacing of the two basic patches varies from 8-30 residues with most exceeding 16 amino acids. That is, at least four, and in the case of VZV up to 18, residues longer than described by the consensus of a classical bi-partite NLS (Marfori et al. 2011; Dingwall & Laskey 1991). On the one hand R4 is, with few exceptions, almost perfectly conserved featuring at least four consecutive basic residues mostly starting with a lysine (Figure 8A, K428, HSV-1 numbering). On the other hand R5, retaining general basic character, diverged but retains an overall positively charged character. All R5s that were evaluated, except for that of equine Herpesvirus-1 (EHV- 1), comprise 5-9 basic residues (of around 20 residues in total) with a preference for lysines in HSV-1/2 and VZV. All sequences show an abundance of prolines (grey shading), multiple serines, threonines and glutamates (PEST sequences) which point to general disorder of this region. (ii) The -subfamily region (Figure 8B) comprises two well conserved basic motifs consisting, with few exceptions (murid herpesvirus 1 and 2 (MuHV-1 and 2)), of two runs of three to four consecutive basic residues divided by a short, variable spacer (8-12 residues). These motifs generally fit the consensus of both mpNLSs and bpNLSs (Table 1 in 1.3.2, Figure 13D). PEST residue clusters can generally be found flanking and within the basic patches. (iii) While the two basic patches are still conserved, the spacer appears further reduced (around 8 residues in all three members examined) in the γ1 subfamily (Figure 8C). Conservation of R4’ is high between representative viruses. This area may represent a mpNLS or a bpNLS with a short 8 residue spacer. In addition, the basic residues are again surrounded by PEST residues indicating general disorder as mentioned for the other regions. Interestingly, the patch of serines and threonines is, due to the

- 78 -

Identifying NLSs of VP1-2 homologues presence of the directly adjacent acidic residues, potentially subject to phosphorylation (Meggio & Pinna 2003). (iv) In the γ2 HVs (Figure 8D) no discernible homology can be found in the most N-terminal region after the USP that exhibits basic characteristics. Despite the lack of conservation class 1 and 2 mpNLS motifs (compare Table 1 and Figure 8D) can be identified in the sequences of bovine Herpesvirus 4 (BoHV- 4) and Herpesvirus Saimiri (HVS), respectively, while the other two sequences cannot be allocated to the described classes and might thus represent an atypical basic NLS. As with the other groups PEST residues are overrepresented in this region.

Figure 8 Alignment of putative NLS regions of the VP1-2 N-terminus of representatives of the different HV sub-families. (A) Position and alignment of the VP1-2 NLS regions of α-HVs (R1 and its sub-regions R2, R4 and R5). The complete highly basic patch was designated R1 in each of the members which corresponds to the first basic region in the linker between the USP and the middle segment. The α-HV R1 region was further subdivided into two basic motifs, R4 and R5. Lysine at position 428 (HSV-1 numbering), the first basic residue of R4, is indicated. (B-D) Alignment of the VP1-2 R1 regions of β- (B), γ1- (C) and γ2- HVs (D). The R1 regions were aligned with the R1 region in the α-HV (see black line in A). The R4 equivalent region in these sub-families was designated R4’ and shows the amino acid sequences I chose as the equivalent region to HSV1.R4. Background colours indicate amino acid groups with similar properties: red (lysine, arginine), yellow (serine, threonine, tyrosine), blue (aspartate, glutamate), green (leucine, valine, isoleucine, methionine, phenylalanine), grey (proline), without background (glycine, histidine, alanine, glutamine, asparagine, cysteine). Amino acids in single letter code. Adapted from (Hennig et al. 2014).

- 79 -

Identifying NLSs of VP1-2 homologues

In conclusion the basic regions were easy to spot in the primary amino acid sequence and aligned reasonably well amongst members of the same sub-family. The described patterns also appeared to be conserved in (parts of) HV sub-families. To derive meaning from the observed differences in organisation it would be intriguing to hypothesise that they evolved as an adaptation to the particular type of host cell these virus groups infect since even HVs infecting different species show similarities.

3.1.2 In silico NLS prediction in the N-terminus of VP1-2 orthologues

To shuttle a nuclear protein into the nucleus cells have assembled an arsenal of nuclear transport receptors which recognise many different signal sequences. As importins not only interact with an NLS but quite often with extensive regions around the NLS of unrelated cargo proteins, multiple, similar transport receptors have evolved to recognise similar NLSs displayed by completely different proteins. Thus finding NLSs can prove a daunting task. To find NLSs in VP1-2 homologues would have probably required me to empirically screen the whole linker sequence between the USP and the body of VP1-2 (and potentially more). To aid my analysis and guide my choice of regions of interest I deployed five different nuclear localisation prediction algorithms to screen for putative signals. These tools can be divided into two groups, which wither consider only the primary sequence (NLSmapper, NLStradamus) or also implement other sequence data including structure and protein composition (Euk mPLoc 2.0, CELLO, NucPredict). For this analysis (summarised in Table 7) I initially analysed the first 1000 residues of VP1-2 homologues. CELLO and NucPredict use a scoring system for the whole polypeptide with higher scores indicating higher confidence that a given sequence localises to the nucleus. Euk-mPloc works similarly but without scoring the input sequences. In contrast, NLSmapper and NLStradamus give an indication whether and where an NLS is present in the input sequence. Not all algorithms indicate which sequence would act as an NLS and some integrate ‘other sequence information’. To account for the different weighting of ‘other sequence information’ I also screened the aligned R1 region (Figure 8), since, judging by the structural conservation of VP1-2, any putative NLS sequence might be found in the first basic patch after the USP. Using the short R1 should allow some of the algorithms to disregard other sequence information which might alter the verdict (e.g. other dominant or cryptic signals such as for secretion). In order to put the findings of in silico nuclear prediction into perspective, these were compared to the truly nuclear SV40 lTag and cytoplasmic β-actin or the HSV-1 VP1-2ΔNLS as positive and negative controls, respectively (Table 6 and Table 7). ‘Positive’ signals (or nuclear localisation) were judged compared to the SV40 signal score (see Table 6 and Table 7, >70% of the score was regarded as positive = green, <40% was regarded as negative = red) Although the algorithms use different analysis methods I identified a NLS in HSV-1 with three (NLSmapper, NLStradamus, Euk mPLoc 2.0) of the five prediction tools if the whole N-terminus was

- 80 -

Identifying NLSs of VP1-2 homologues considered. This could be improved to four out of five algorithms using only R1 (CELLO). The prediction of NLSmapper and NLStradamus even picked up the previously confirmed HSV-1 NLS but additionally indicated that R1 might actually comprise a bpNLS. The other α-HVs (VZV, PRV, BHV-1) scored positive up to two times. The VZV VP1-2 equivalent region scored low in CELLO and NucPredict or was not predicted to contain any putative NLS sequence in R1. The predicted NLS in the VZV VP1-2 homologue was suggested to lie further downstream, outside the linker region. Taking into account VP1-2 structural conservation and recent evidence that the central region of VP1-2, where the NLS was predicted, forms a ordered structure (Scrima et al. 2015), it would be of no relevance for my studies. Interestingly, the BHV-1 homologue was predicted NLSmapper to comprise an NLS even though the linker region lacks the first lysine (similar to VZV). A possible explanation why the putative VZV NLS was not picked up could lie in the unusually long spacer between the two basic patches within the VZV.R1 (30 residues compared to 8 for BHV-1) and the presence of a KR motif four residues upstream of the BHV-1 R4. probably because of the long linker and lack of a lysine in the first basic patch (Figure 8). The PRV sequence at least twice scored lowest (NucPredict with 0.35 and CELLO with 1.054), although it was previously confirmed to contain two functional NLSs (Mohl et al. 2009) and adheres to the consensus of a mpNLS better than the corresponding region of VZV (Figure 8, see KRRR motif). The sequences of the other HV sub-families were more heavily linked with nuclear localisation. As such HCMV was predicted to localise to the nucleus with all algorithms except for NLSmapper. In fact, the confirmed bpNLS (Brock et al. 2013) was correctly picked up by NLStradamus. The sequences of the EBV, HHV-8 and HVS homologues were also consistently predicted to contain a NLS (first 1000 residues). Interestingly, HHV-8.R8, which did not look like a NLS at first glance due to presence of so few basic residues, was also spotted with NLSmapper and NLStradamus. The difference between the results of the long and short regions highlights the varying weighting each algorithm gave to whole sequence. NucPredict, CELLO and Euk-mPloc 2.0 look more at overall protein composition and may thus be more useful to analyse uncharacterised proteins. NLSmapper and NLStradamus appeared to search for NLSs in the primary amino acid sequence, which would probably make them useful if one knows the protein structure and thus where importin-accessible regions lie. However, these algorithms have their limitations. For once the analysis of the VP1-2ΔNLS sequence only had a minor effect on the verdict using four out of five tools (only NLSmapper predicted lack of a NLS). Moreover, it would be very difficult to spot unconventional NLSs with these algorithms and to consider changes to cellular physiology that happen during events such as viral infection. In conclusion I was able to narrow down the search for NLSs to mostly R1 (Figure 8) although not all sequences were consistently identified (e.g. VZV). After this exercise I was confident that I was looking at the correct region (R1) to test for NLS function. The data obtained by this in silico analysis, however,

- 81 -

Identifying NLSs of VP1-2 homologues cannot substitute for empirical (manual) analysis of the primary sequence and required experimental verification.

Table 6 NLS prediction of two proteins that contain a confirmed classical mpNLS (SV40) or a confirmed bpNLS (nucleoplasmin).

Euk-mPLoc NLSmapper NLS Accession Virus NucPredict CELLO 2.0 NLS score tradamus

P03070 SV40 T-ag 0.88 2.601 Nuclear Mono 16 127-130

P60709 β-actin 0.09 0.18 Cytoplasmic None found None found

- 82 -

Identifying NLSs of VP1-2 homologues

Table 7 Overview of NLS predictions in the N-terminus of VP1-2 homologues using five different, freely available NLS prediction algorithms. Protein sequences were entered in FASTA format into the online tool which predicts localisation either associated with a likelihood score (CELLO, NucPredict; the higher the score the more likely the protein may be found in the nucleus), or showing the region containing the putative NLS (NLStradamus, NLSmapper) or giving an overall verdict (Euk-mPLOC 2.0, CELLO). As control the cytoplasmic VP1-2ΔNLS mutant (Abaitua & O’Hare 2008) was included in this analysis. Red font = <40% of SV40 lTag score in NucPredict and CELLO only (NucPredict = 0.88; CELLO = 2.601), green font >70% of SV40 lTag score, orange font between 40-70% of SV40 lTag score. For Euk-mPloc 2.0, NLSmapper and NLStradamus green font indicates a positive result (nuclear or NLS), red font indicates a negative result (e.g. extracellular). Algorithms based on Yu et al. (2006); Nguyen Ba et al. (2009); Chou & Shen (2010); Kosugi et al. (2008); Kosugi, Hasebe, Matsumura, et al. (2009); Kosugi, Hasebe, Tomita, et al. (2009); Brameier et al. (2007).

NLSmapper NucPredict CELLO Euk-mPLoc 2.0 NLS tradamus Accession Virus aa 1-1000 R1 (or R8) aa aa aa 1-1000 R1 (or R8) R1 (or R8) aa 1-1000 R1 (or R8) NLS Score NLS Score 1-1000 1-1000

HSV-1 0,51 0,46 1,428 2,951 Nuclear, cytoplasm extracellular mono, bi 8, 7.9 mono, bi 8, 7.9 425-467 ABI63498.1 HSV-1 0,38 0,26 1,32 2,089 Nuclear, cytoplasm extracellular none found 442-461 NLS

mono (not A6XEB1 VZV 0,5 0,34 0,841 2,024 Nuclear extracellular 9 none found none found R1) Q5PPB8 PRV 0,35 N/D 1,054 N/D Extracellular N/D mono, bi 9, 5.5 N/D N/D 278-297 Q65553 BHV-1 0,43 N/D 1,39 N/D Extracellular N/D bi 5,6 N/D N/D N/D

P16785 HCMV 0,77 0,63 2,216 2.701 Nuclear extracellular, nuclear none found 281-304

K9UT36 EBV 0,84 0,68 3,485 2.866 Extracellular, nuclear nuclear mono, bi 5, 5.2 mono 5 411-432

0.38 (R1) 2.45 (R1) nuclear (R1) none found Q2HR64 HHV-8 0,8 2,723 Nuclear mono 8 304-322 0.57 (R8) 2.336 (R8) nuclear (R8) mono (R8) 8 (R8) 0.16 (R1) 2.26 (R1) nuclear (R1) none found Q01056 HVS 0,72 3,516 Nuclear bi 8,2 258-308 0.32 (R8) 1.896 (R8) nuclear (R8) bi (R8) 8.2 (R8)

- 83 -

Identifying NLSs of VP1-2 homologues

3.1.3 Analysis of NLS functionality

To test whether the predicted NLSs are functional motifs, several approaches could be employed. (i) Nuclear localisation could be analysed using VP1-2 homologues from representative HV sub-family members. This would be considered the most physiological approach. (ii) Putative NLSs could be screened using a heterologous protein, such as GFP or similar. This would allow looking at isolated regions which could be useful since it would effectively remove NLS presentation issues. (iii) NLS regions from analogous positions of representatives from other HV sub-families could be cloned into the HSV-1 VP1-2 protein since domain structure appears to be conserved across all VP1-2 homologues (see Figure 3). This would allow screening regions of interest in the context of a conserved protein at an analogous position. (i) Since cloning the full length VP1-2 homologues from the other sub-families could be problematic the first approach could take a long time to complete. This would have to be followed up by more detailed analysis using mutations to identify the actual NLS. (ii) Our group had already successfully used the β-gal system to identify the HSV-1 NLS (Abaitua & O’Hare 2008). Thus it seemed reasonable to continue with this system and clone any predicted putative NLS sequence into the N-terminus of it. However other heterologous proteins, including GFP, that could also be suitable for screening putative NLSs, should be considered. These systems are useful since they isolate the test region at the protein terminus it is fused to, although this may lead to false positive results. Any positive signal still has to be confirmed with another technique (see iii). (iii) There are multiple reasons why it could be feasible to study the homologous R1s in the context of HSV-1.VP1-2. (a) Our lab has already cloned and worked with HSV-1 VP1-2 and the HSV-1 BACMID system. (b) The localisation of VP1-2 has already been well characterised (Abaitua & O’Hare 2008; D S McNabb & Courtney 1992; Möhl et al. 2009). (c) NLSs usually present as linear epitopes, which is the reason why NLSs tend to be transferrable (if this is their sole function) and only require a cellular transport factors to be present (e.g. importin α subtypes). (d) The linker region connecting the USP and middle segment is not well conserved (while the USP and middle segment are) and most importantly not predicted to contain secondary structure that could be disrupted by inserting a foreign sequence. (e) The VP1-2 homologues show high levels of conservation across independent areas, and especially within the putative NLS regions. The areas surrounding the putative NLS all contain multiple helix/sheet breaking residues indicating a similarly disordered character as the HSV-1 part. Assuming that nuclear docking was the only function of the 70 residue patch around the NLS, this part of the linker between USP and middle segment can be swapped directly (Figure 14A) to allow functional analysis. Domain or even complete protein transfer between related viruses has been done before and can be a useful tool to study protein function (Kuhn et al. 2008). (f) Finally, for the HSV-1 protein our

- 84 -

Identifying NLSs of VP1-2 homologues group has already shown that deletion of part of the NLS allows for normal virion formation (including recruitment of other tegument proteins) and replication precluding a structural function of this motif in VP1-2 and in the context of viral capsids. Thus I used two complementing approaches mentioned in (ii) and (iii). Initially, to screen for any motives that contain NLS function, a non-quantitative screen utilising the large bacterial protein β- galactosidase (β-gal) was deployed. For this, regions of interest were cloned in frame into the N- terminus of β-gal. This approach was designed to find any putative NLS motifs regardless of their strength or whether they could function in physiological context, since they will be presented as linear epitopes. All fusion constructs were screened for their ability to localise β-gal to the nucleus. The second approach was designed to assess NLS function in a more physiological setting and to study NLS function/efficiency qualitatively and quantitatively. For this, the extended NLS regions (R1) from representative VP1-2 homologues from all HV sub-families were inserted into HSV-1 VP1-21-1875 (NT6) which would mimic a more physiological context of the VP1-2 NLS.

3.2 Results

3.2.1 Development of the -galactosidase nuclear translocation assay

In order to evaluate the NLS activity of a variety of regions of interest it was important to test their capacity to translocate a cytoplasmic protein to the nucleus. There were three options for screening NLS activity in VP1-2 homologues. For an initial screen I chose to pursue a heterologous protein translocation assay to find functional NLS motifs. Initially, I wanted to see which was the best screening protein for my studies (i.e. not nuclear without NLS and nuclear with NLS and easy to detect). For this I chose to investigate, GFP and -gal fused to known HSV-1 NLSs (Abaitua & O’Hare 2008). Although these screening vectors do not show the same structural constraints as VP1-2, where the putative NLS presumably lies between two folded domains (USP and middle segment, Figure 3), they are fully folded proteins which are not disrupted by adding short or even long sequence tags. In fact, any sequence added is likely solvent exposed and thus freely accessible to the nuclear transport machinery which I assumed would be the case in the VP1-2 linker. This approach was not designed to score functional efficiency of test NLSs but to find any sequence able to confer nuclear localisation to a heterologous protein, and it will be complemented by a second study of NLS activity within HSV-1 VP1-2.

- 85 -

Identifying NLSs of VP1-2 homologues

Figure 9 Development of the nuclear translocation assay; comparing GFP and β-gal NLS fusions. (A) Overview of the sets of constructs planned for this thesis. For β-gal and GFP the NLSs were fused in frame into its N- terminal. (B, C) Representative images of fixed HuH7 cells show steady state localisation of (B) GFP (I), GFP-R4 (II) and GFP- R2 (III) or (C) -gal (IV) and β-gal fusions with SV40 lTag NLS (V), R4 (VI) and R2 (VII). Cells were fixed 24h post-transfection (B, C) and stained with anti β-gal antibody (C). Nuclei were counterstained with DAPI. R4 and R2 are HSV-1 specific sequences (see Figure 8). Cells were imaged at 63x magnification.

For the purpose of this particular analysis, NLS regions R2 and R4 were fused in frame to GFP and β-gal (see Figure 9). GFP itself (panel I) showed both nuclear and cytoplasmic signal but was enriched in the nucleus even in the absence of any exogenous NLS. Upon fusion of R4, the smaller of the NLS regions, to GFP, the fusion protein appeared to enrich in the nucleus not much beyond what was seen for GFP alone as judged by remaining cytoplasmic signal (Figure 9B, panel II). When the longer

- 86 -

Identifying NLSs of VP1-2 homologues

R2 region containing the second basic patch (i.e. R2 = R4 + R5) was fused to GFP, the fusion protein now almost exclusively localised to the nucleus (Figure 9B, panel III) implying there are additional signals that either enhance nuclear translocation or that retain the protein in the nucleus. Upon closer inspection, R2-GFP appeared to accumulate in nucleoli as the regions of strong green signal overlapped with areas of low DNA density. The lack of DAPI signal indicates a lack of DNA in that region which is a characteristic feature of protein dense nucleoli. In silico analysis (Nucleolar localisation sequence detector, University of Dundee) of R1 predicted a nucleolar localisation signal in the C-terminus of the NLS (i.e. R5) which could be causing this sub-nuclear compartmentalisation (Scott et al. 2010). β-gal exhibited a mainly cytoplasmic diffuse pattern of localisation (Figure 9C, panel IV) with very little staining observed in the nucleus. However, the strength of the nuclear signal varied from experiment to experiment but was generally much weaker than in the cytoplasm. If the strong experimentally verified NLS of the SV40 lTag was fused N-terminally to β-gal the fusion protein localised exclusively to the nucleus (Figure 9C, panel V), validating the assay format. Fusing R4 to β-gal (Figure 9C, panel VI) showed a similar effect on localisation as the SV40 NLS, which was not enhanced by extending the NLS to include the second basic patch (Figure 9C, panel VII). Intriguingly, R2.β-gal did not recapitulate the nucleolar accumulation found with GFP-R2Actually, R4.β-gal appeared to be excluded from distinct weakly DAPI-stained sub-nuclear compartments. GFP is a small fluorescent protein making it useful for live cell imaging. However, its relatively small size of 27kDa allows it to enter the nucleus without the need of an NLS (Figure 9B, panel I). On the other hand the bacterial protein β-gal is large (130kDa) enough to be excluded from nuclei (Figure 9C, panel IV) and does not contain any putative NLS or NESs making it a good candidate cargo. The latter fact makes it ideal to find even weak NLSs as once nuclear, it remains there unless nuclear integrity is disrupted. The fact that R4 NLS activity might have been missed using GFP as a heterologous transport cargo lead me to focus on β-gal as a screening tool to find NLSs, since there would be a higher chance to identify even weak signals.

3.2.2 Analysis of NLS function of the VP1-2 R4 regions using the -gal system1

After establishing the screening system I could now think about which regions to test first. Due to the striking conservation in the -HVs and the relatively strong conservation of basic residues in the other sub-families, I initially analysed a homologous region (R4 for α-HVs and R4’ for the other HV sub- families as shown in Figure 8) around the core basic residues in representative human viruses from all

1 Figures of this sub-chapter were adapted from (Hennig et al. 2014)

- 87 -

Identifying NLSs of VP1-2 homologues three sub-families: VZV (), HCMV () EBV (1) and HHV-8 (2). These regions were chosen due to their position within VP1-2 (first basic patch after the USP domain) and cloned in-frame into the N-terminus of the -gal ORF. For analysis of localisation constructs were transfected into HuH7 cells and their localisation examined at steady state after 24 h of expression by indirect immunofluorescence to β-gal (see Figure 10).

Figure 10 Localisation of β-gal fusion proteins comprising R4 homologous regions from different HV homologues. HuH7 cells were fixed 24h post-transfection. Localisation of fusion protein was detected using anti β-gal antibody and nuclei were counterstained with DAPI. Representative images are shown. Each panel consists of two images, top and bottom, showing the FITC and the DAPI channel, respectively.

The control constructs (Figure 10, panels I-III) exhibited exclusively nuclear localisation for NLS sequences of SV40 lTag and R4 (panels II and III, respectively) and mostly cytoplasmic accumulation for β-gal alone (panel I). Unexpectedly, the highly conserved VZV.R4 fusion shows mainly cytoplasmic distribution (Figure 10, panel IV) similar to that of the negative β-gal control (panel I). This might indicate that either the VZV homologue does not contain a functional NLS or that the NLS may lay downstream or encompasses a larger region. This result also raised the question whether this would be similar for the other representatives of the - and -HVs. Expectedly, the HCMV.R4’ and EBV.R4’ regions containing both short basic patches showed strong activity exhibiting almost exclusively

- 88 -

Identifying NLSs of VP1-2 homologues nuclear localisation which parallels the activity of the HSV-1 peptide (Figure 10, panel V and VI, respectively). To cover all the different types of linker regions present in the HV sub-families, it was necessary to include a 2 representative, namely HHV-8, since the sequence barely resembles that of any other sub-family (Figure 8D). HHV-8.R4’ lacks N-terminal lysines but contains several arginines clusters of no more than two consecutive residues, thus, perhaps as anticipated, it did not enable nuclear translocation of -gal beyond control levels (Figure 10, panel VII). As for VZV, the NLS activity might be absent from this region or still be found further downstream or be part of a multi-partite signal.

3.2.3 Analysis of NLS function of the extended VP1-2 R1 regions using the β-gal system

To expand upon the study of R4s and to clarify the lack of NLS activity in VZV and HHV-8 the test region was expanded to include all basic residues within an approximately 70 residue stretch around R4 (designated R1, Figure 8). For this, HA-tagged R1s were cloned in-frame into the N-terminus of β- gal and localisation was analysed by indirect immunofluorescence to the HA tag. Consistent with the results from R4, HSV-1.R1 showed strong nuclear accumulation (Figure 11, panel III) whereas HA-β-gal accumulated in the cytoplasm (Figure 11, panel I). I noted marginally increased nuclear signal with this control compared to β-gal lacking the HA-tag but the dynamic range between negative and positive control (HA-β-gal vs HA-NLS-β-gal) was still large enough to be easily analysed. A further control using a cellular NLS (Myopodin; MyoNLS) inserted within HSV-1.R1 in place of the first basic patch in R4 (Figure 11, panel II) also showed nuclear staining. Since β-gal cannot exit the nucleus once it entered, it was thus difficult to judge if the strong MyoNLS could improve upon R4 activity, underscoring the limits of this nuclear transport assay system.

- 89 -

Identifying NLSs of VP1-2 homologues

Figure 11 Localisation of β-gal fusion proteins featuring R1s from HV VP1-2 homologues. HuH7 cells were fixed 24h post-transfection and the N-terminal HA-tag was used to detect the fusion protein. As an additional control the Myopodin NLS was cloned into R1 in place of the first basic patch of R1 (Myo.R1). Nuclei were counterstained with DAPI.

Intriguingly, the expanded region of VZV showed strong nuclear staining (Figure 11, panel IV), which indicates the NLS might be bi-partite. Consistent with the result from the analysis of R4s, the CMV and EBV fusions confirmed that the R1 region of these representative HVs contains a NLS (Figure 11, panel V and VI, respectively). In contrast to this, extending the HHV-8 test region downstream of R4 to include additional basic residues did not lead to detectable nuclear localisation function in this assay (Figure 11, panel VII). Thus the first large basic patch downstream of the USP in HHV-8 VP1-2 does not contain a NLS. This function may not be present in this region but this requires further experimental validation. These results lead to the conclusion that all respective VP1-2 regions contain a NLS if these comprise a run of at least three basic residues with an N-terminal lysine or two patches with one consecutive run of at least three basics and additional lysines further downstream (compare Figure 8). The area in HHV-8 showed an overall basic character but did not satisfy these criteria and thus probably

- 90 -

Identifying NLSs of VP1-2 homologues not surprisingly did not confer nuclear localisation to β-gal.Further analysis of NLS function in HHV-8 VP1-2 using the β-gal system Since all VP1-2 homologues except HHV-8 were found to contain a NLS in R1 I wanted to check whether this function was present elsewhere in the linker region. Thus I extended the analysis to to a region further downstream of R4 (Figure 12B). The linker region indeed contained areas comprising further arginines or lysines (Figure 12A, red vertical lines) and two, R5 and R9, were chosen for analysis. In fact R5 was predicted by NLStradamus and R9 conforms loosely to the consensus of a class 3/4 mpNLS (KRKF). Although HHV-8 showed low conservation, the region designated R8 aligned comparatively well amongst γ2 members if the heavily acidic patch was used as a point of reference (Figure 12C). At least one other γ2-HV (Figure 12C, BoHV-4) actually comprised a stretch of four basics with an N-terminal lysine resembling the class 1 NLS consensus. Surprisingly, HHV-8.R8 also bore some limited resemblance to the γ1-HV.R1 (Figure 12D) when the region was aligned based on the homology of the short HHV-8 basic and acidic patches. The negative controls β-gal (Figure 12E, panel I) and HHV-8.R4-β-gal (panel II) again showed no nuclear localisation. Since HHV-8.R1 was already known to not confer nuclear localisation to β-gal (panel IV) it was maybe not surprising that HHV8.R5 was also functionally inert in this assay (panel III) although it includes two more basic residues that were not included in R1 (see Figure 12A). Unexpectedly, and despite the presence of only five basic residues and generally highly acidic character of HHV-8.R8, this sequence conferred nuclear localisation to β-gal (Figure 12E, panel V). The actual NLS, however, precedes the acidic patch and resides in the four basic residues of R9 (Figure 12E, panel VI) implying that this motif is a mpNLS These data taken together with the results on the other HV motifs I conclude that likely all HVs contain a NLS in the linker region between USP and middle domain which could function similarly during entry of HV capsids.

- 91 -

Identifying NLSs of VP1-2 homologues

Figure 12 Analysis of a basic region in HHV-8 VP1-2 downstream of R1. (A) Schematic overview of the location of basic residues in the non-conserved linker region of HHV-8 VP1-2. The linker is expanded below and the red vertical lines indicate the relative position of basic residues within the linker region. Sub- regions are indicated by black lines. (B) Alignment of γ2 HV VP1-2 region around R1 to compare to the improved alignment in C. R4’ and R5 = black line, R1=red dotted line. (C) Alignment of an acidic region found in the γ2 representatives shows upstream conservation of a short basic patch which conforms loosely to the NLS consensus. (D) Alignment of HHV-8 R8 with R1s of γ1-HVs, which show some conservation around the central KRKK motif including a downstream polar-acidic patch. R9 = black line, R8 = red dotted line. (A-D) Amino acids in single letter code with same colour coding as in Figure 8. (E) Representative images showing the localisation of N-terminal β-gal fusions of the in (A) depicted regions detected with anti-HA antibody in HuH7 cells.

3.2.4 Analysis of determinants of NLS function of HSV-1 VP1-2 using the β-gal system

The striking difference in activity between the HSV-1.R4 and VZV.R4 was a reason to look more closely at the determinants that govern NLS function of this region. Although these two α-HVs contained a NLS in R1 it came as a surprise that VZV.R4, despite showing a very high degree of conservation, was an exception. Thus, I first examined the sequence around R4 in more detail. Although they look very similar (Figure 8), there are three main differences: (i) There is a run of five basic residues in at least two VZV strains compared to the four basics found in most α- and β-HVs. (ii)

- 92 -

Identifying NLSs of VP1-2 homologues

VZV.R4 lacks a lysine at the first position of the basic run in R4 compared to the HSV-1 sequence (Figure 8A, equivalent of HSV-1 K428). This also deviates from the consensus of mp or bpNLSs (see Figure 13D and Table 1). (iii) The gap between the first and second basic patch counts 30 residues compared to 16 for HSV-1 which may affect the utility of a bpNLS but should not affect mono-partite signals. To examine these peculiarities further I introduced mutations into the previously tested regions (R1 and R4) of both HSV-1 and VZV and checked whether this affected the localisation of the respective β-gal fusion proteins. Representative images of the controls for this assay are depicted in Figure 13A which show nuclear localisation for HSV-1.R4 and HSV-1.R1 (Figure 13, panels II and III, respectively) while VZV.R4 and β-gal alone remained cytoplasmic (panel IV and I, respectively). The R>P substitution of the first arginine from VZV.R4 confirmed that the run of five basic residues was not the culprit for lack of function of this region (Figure 13B, panel VI). Vice versa, the P427R substitution in HSV-1 had no effect on function of the HSV-1.R4 sequence (Figure 13B, panel V). This also indicates that the reasonably well conserved proline plays no direct role in recruitment of nuclear transport factors in this assay. The consensus of mpNLSs (Figure 13D, top two lines) specifies a lysine at the most N-terminal position of the basic patch. To see whether the lack of a lysine did indeed play a role in NLS function, a second mutation was introduced into both R4s. The K428R substitution in HSV-1 completely abolished nuclear localisation consistent with the consensus requirements of a mpNLS (Figure 13B, panel VII). Vice versa, introducing a lysine at the analogous position (the second arginine) in VZV.R4 (Figure 13B, panel VIII) conferred nuclear localisation to -gal. This is not further enhanced by the double mutation, RR>PK, in VZV.R4 (Figure 13B, panel IX). These results indicate that the VZV structure likely does not encompass a mpNLS in R4 due to the lack of a lysine at position 432 (VZV numbering). The above mutations looked at R4 in isolation but since the second basic patch was conserved in its position (i.e. downstream of R4) it was reasonable to assume it would aid R4 function since both basic patches contribute to transport factor binding. Thus the same mutations were introduced into R1-β-gal fusions. Interestingly, the K428R mutation in HSV-1.R1 (Figure 13C, panel XI) did not recapitulate the deleterious phenotype of the same change in the context of HSV-1.R4 (Figure 13B, panel VII, HSV-1.R1.K428R showed NLS activity but R4.K428R did not). This corroborated the finding that the VZV.R1 harbours NLS activity as part of R1 but not R4 (Figure 13A/C, panels IV and X) since it also lacks a leader lysine. To further test whether the first basic residue would at all be required for NLS activity as part of R1 a non-conservative mutation, K428A, was introduced into HSV-1.R1, which should sufficiently destabilise importin-NLS interaction. Expectedly, this change abolished nuclear localisation completely (Figure 13C, panel XII) confirming that R4 has a dominant role in conferring NLS activity. Conversely, this serves as an additional control showing that R5 alone is not a functional NLS

- 93 -

Identifying NLSs of VP1-2 homologues

(Figure 13C, panel XIII) but R4 is (Figure 13C, panel XIV). Taken together, this indicates that the second basic patch of HSV-1.R1 contributes to NLS function in HSV-1, although R4 may function as a mpNLS in isolation. Expanding upon previous findings I hypothesise that this region likely functions as a bi-partite motif during viral replication as well.

- 94 -

Identifying NLSs of VP1-2 homologues

Figure 13 Effect of point mutations in the VP1-2 NLS of HSV-1 and VZV on localisation of -gal fusion proteins. Mutations were introduced into the HSV-1 NLS region and the analogous residues of the VZV NLS. (A) Localisation of control fusion constructs described before. (B, C) Mutational study of the R4 region (B) and of the R1 region (C) of HSV-1 and VZV.

- 95 -

Identifying NLSs of VP1-2 homologues

Mutated NLSs were cloned in frame into the -gal ORF (also contained N-terminal HA-tag). Localisation was detected in HuH7 cells using anti-HA antibody 24h after transfection. Representative images are shown for each construct (incl. DAPI). (D) The consensus of classical a mpNLS and bpNLS that would conventionally employ importin- and  mediated nuclear import pathway (see Table 1). This is compared to the mpNLS of HSV-1 (in R4) and the potential bpNLS variants that could enable nuclear transport in HSV-1 and VZV. Residues designated P1, P2, etc. show positions in a mpNLS or the second basic patch of a bpNLS. Positions designated P1’ and P2’ show the first (upstream) part of a bpNLS. X = any amino acid with numbers in subscript indicating number of residues. Red letters are important consensus residues. Green residues match the consensus and blue residues show deviation from the consensus.

- 96 -

Identifying NLSs of VP1-2 homologues

3.2.5 Studying NLS function in the context of HSV 1 VP1-2

To complement and confirm the β-gal data I designed a strategy to insert all R1s into HSV-1 VP1-2 (Figure 14A) which enabled me to easily screen multiple sequences. For this, the VP1-2 DNA sequence was modified by inserting two restriction sites near the NLS without changing the protein sequence. These sites were subsequently used to insert any foreign sequence (such as the R1s from VP1-2 homologues). To create an easy to interpret nuclear transport assay system a few things required my consideration. Firstly, the positive control should preferably localise exclusively to the nucleus and, secondly, this localisation should optimally depend on the presence of a NLS only. However, the localisation pattern of full length VP1-2 is heterogeneous showing cytoplasmic and nuclear localisation likely caused by the presence of multiple domains with separate but presumably, at least in the context of transfection, competing functions (D S McNabb & Courtney 1992; Abaitua & O’Hare 2008; Möhl et al. 2009; Brock et al. 2013; Schipke et al. 2012). Thus in order to create a robust and easy to analyse system to study the nuclear localisation function of the homologous R1s in the context of HSV-1 it was necessary to truncate VP1-2. The truncated VP1-21-1875 protein, designated NT6 (N-terminal product 6), which had the C-terminal 40% removed, is too large to enter the nucleus by simple diffusion and showed strong nuclear localisation in the majority of cells (Figure 15, panel I and B). This was entirely dependent on the presence of the NLS in R1 (Figure 15, panel II and B) and not due to passive diffusion (too large) or presence of a second motif with similar activity, since deletion of the first basic patch almost completely abolished nuclear localisation. Using this system it was possible to test for NLS function of R1s in a more physiological context.

3.2.6 Expression of chimeric VP1-21-1875

Due to the size of full-length VP1-2 it also is not expressed too well in transient transfections. C- terminal truncation should improve expression and stability of the protein (VP1-21-1875) I wanted to check whether NT6 can be recovered from cells intact, since it was suggested that VP1-2 could be cleaved during its life span (Kattenhorn et al. 2005; Jovasevic et al. 2008; Fuchs et al. 2004). Shorter, cleaved forms of NT6 could potentially migrate into the nucleus freely, which could obscure the data. All of the test NT6s expressed sufficient amounts of the protein in transfected cells as detected by Western Blot analysis (Figure 14). Breakdown products of NT6 could sometimes be found (compare lanes 2 and 11 in Figure 14B) but detection of full length product varied only marginally between the different constructs in the same experiment. The occasionally observed breakdown could relate to sample handling and may be minimised by omitting the sonication step to fragment cellular DNA (e.g.

- 97 -

Identifying NLSs of VP1-2 homologues treat lysate with DNase prior to adding SDS buffer, dilute sample or shear DNA). Since NT6 was tagged at the N-terminus, the observed breakdown products between 40-58 kDa (Figure 14B, left) could obscure the localisation pattern since small proteins (usually <40kDa) can enter the nucleus through diffusion. This should be taken into account for the later described localisation analysis.

Figure 14 (A) Cloning strategy to create chimeric NT6 derivatives and their expression in HuH7 cells. Unique BamHI and MluI sites in the 1875 residue NT6 portion of VP1-2 were used to insert the homologous test regions (R1s, in red) of representatives of HV sub-families via an intermediate vector containing a chemically engineered BamHI- MluI fragment containing two additional unique restriction sites, XhoI and NheI (red arrows), which were used for further cloning. All chemically synthesised R1s (and HSV-1 mutants) were inserted via XhoI and NheI restriction sites into the BamHI-MluI carrying vector and subsequently transferred to NT6 via ligation between BamHI and MluI. Colour coding indicates sequence homology between different regions compared to HSV-1. (B) Western blots showing expression of NT6 in HuH7 cells transfected for 24h. NT6 protein in cell-equivalent fractions was detected using anti-V5 antibody.

3.2.7 Analysis of localisation of mutant NT6

After confirming expression of the chimeric and mutant NT6s in transfected cells the next step was to qualitatively and quantitatively analyse protein localisation. For qualitative analysis HuH7 cells were utilised and representative images of localisation patterns are shown in Figure 15B.To minimise localisation artefacts NT6 cellular distribution was examined in three different cell lines and more than 100 cells were scored. To simplify analysis the observed localisation phenotypes were allocated to three categories (Figure 15A, HuH7 cells depicted); nuclear only + nuclear > cytoplasmic (cat I, Figure

- 98 -

Identifying NLSs of VP1-2 homologues

15A, panels I and II), evenly distributed signal (Figure 15A, cat II, panel III) and cytoplasmic > nuclear (Figure 15A, cat III, panel IV). As previously mentioned, the controls of the nuclear transport assay exhibited a large dynamic range. NT6 showed strong nuclear localisation with the majority of cells showing only nuclear staining (panel I, Figure 15B/C). This was entirely dependent on the presence of the NLS in R1 (Figure 15B/C, panel II). With these robust benchmarks nuclear localisation of NT6 derivatives could be tested by indirect immunofluorescence in transfected cells (Figure 15B, panels III-XI). More than 75% of HuH7 cells transfected with NT6 harbouring the VZV.R1 sequence show increased nuclear localisation compared to NT6ΔNLS (Figure 15A, panel II and III; B). However, VZV.R1, which shows the highest level of conservation compared to HSV-1 (Figure 8), caused a 50% reduction of cells that show only nuclear or mostly nuclear staining (B)). In fact, almost no cells were found to contain only nuclear staining. This conclusion was corroborated by results obtained using the NT6.K428R mutant (Figure 15B, panel IX; B) which recapitulates the drop in nuclear localisation found for NT6.VZV.R1 (B). Furthermore, the introduction of the K428A mutation (Figure 15B, panel X) almost completely abolished nuclear translocation as was previously found for the β-gal fusion protein. The role of the second basic patch was initially hinted by the failure of VZV.R4 to translocate β-gal to the nucleus so I was interested in its role in the context of VP1-2. In agreement with previous results deletion of R5 still allowed nuclear translocation of NT6 (Figure 15B, panel VIII) although NLS efficiency dropped to a level comparable to the VZV sequence (B). To see whether NLS activity could be compromised by insertion of a different NLS in place of the basic patch in R4 a cellular mpNLS (from Myopodin) was inserted. This construct localised to the nucleus with similar efficiency as wild-type NT6 (Figure 15B, panel XI, B), indicating that the combination of a different, strong NLS and R5 did not interfere with function of R1 although R5 was required for full function previously. Taken together these data imply that R4 has the dominant role in facilitating nuclear transport of VP1-2 and R5 only contributes to this since deletion of R5 had only a modest effect on whereas K428A abolished nuclear transport. In addition the lysine (K428) plays a crucial role for full NLS function in the context of NT6 and probably VP1-2. This leads to the conclusion that the NLS in HSV-1 can function as a weak mpNLS and a stronger bpNLS and that for full activity integrity of HSV-1.R1 is required. Notwithstanding the dramatic changes in sequence and organisation of R1s in the β and γ1 sub- families, nevertheless the CMV.R1 (Figure 15B, panel IV) and EBV.R1 (Figure 15B, panel V) regions showed strong NLS activity which looked at times more efficient at translocating NT6 to the nuclear compartment than their wild-type counterpart. The γ2-HV regions from HHV-8 recapitulated what had

- 99 -

Identifying NLSs of VP1-2 homologues already been found for the β-gal fusions. HHV-8.R1 showed no NLS activity above that of NT6ΔNLS in HuH7 cells (Figure 15B/C, panel VI). In contrast, HHV-8.R8 showed activity comparable to the wild-type sequence (Figure 15B, panel VII). This shows that NLS activity is transferrable between the different VP1-2 homologues and that the structure of NT6 probably was not affected by insertion of 70 foreign residues. Next I examined the effect of mutations within the HSV-1.R1 region. These generally affected nuclear transport of NT6. Expectedly, the K428A substitution (Figure 15B/C, panel X) essentially abolished nuclear translocation, which was in agreement with the previous β-gal study. If the conservative substitution K428R was introduced into NT6, the NLS remained functional (Figure 15B/C, panel IX) although at modestly reduced level. In theory, the stronger a NLS binds to its importin receptor the more efficiently it traverses the NPC (Hodel et al. 2006) implying that the K428R mutation destabilised the importin-NLS interaction. The most prominent reduction in utility of the K428R NLS was observed in HEK293 cells. Interestingly, the K428R mutant NLS still functioned more efficiently than VZV.R1 (Figure 15C) implying that other features, probably residing in the spacer, R5 and/or the surrounding area contribute to function (which are HSV-1 specific and only work fully in conjunction with the HSV-1 NT6). Deletion of R5 modestly affected NT6 localisation in HuH7 cells (Figure 15B/C, panel VIII) whereas it had a more pronounced effect on NT6 localisation in HEK293, although NT6ΔR5 still showed some nuclear localisation above the level of NT6ΔNLS (Figure 15C). These data were in agreement with previously obtained results for β-gal fusion protein in thatR4 is the dominant feature for nuclear localisation in NT6 and probably VP1-2, while R5 only contributes accessory functions. To lower the chance of localisation artefacts due to the choice of cells, two further easily transfectable cell lines were chosen (COS-1 and HEK293) to score localisation of some constructs with lowered nuclear translocation efficiency as described before. In quantitative terms the results were largely consistent with the data obtained for localisation in HuH7 cells although it appeared that in HEK293 cells nuclear transport was more tightly regulated (Figure 15C, right graph). This could have been caused by differences in availability of nuclear transport factors in the different cell types.

- 100 -

Identifying NLSs of VP1-2 homologues

- 101 -

Identifying NLSs of VP1-2 homologues

Figure 15 Localisation of NT6 proteins (described in Figure 14) in HuH7 cells. (A-C) The localisation of transiently expressed chimeric or HSV-1 mutant NT6 was detected using anti-V5 antibody. Four major localisation patterns (A) were observed which were categorised into three categories; category I includes all nuclear only and nuclear > cytoplasmic events, category II includes cells showing even distribution of signal (nuclear = cytoplasmic), category III includes all cells showing nuclear < cytoplasmic staining for NT6. The white arrowhead indicates the cell of interest. Representative images of the main phenotypes observed in HuH7 for the different constructs are shown in B. The localisation pattern was scored manually (for categories see A) in more than 100 HuH7 or HEK293 cells for NT6 mutants which showed intermediate localisation phenotypes (C).

- 102 -

Identifying NLSs of VP1-2 homologues

3.3 Discussion

Work presented in this chapter has led to the discovery of a NLS in approximately the same position in VP1-2 homologues of representatives from all HV sub-families, although the organisation of the region varied considerably between sub-families. NLS activity was confirmed in a heterologous system and in the context of HSV-1 VP1-2 in a variety of cell lines. Furthermore, I was able to show that the R1 region of HSV-1 and VZV, and probably all other α-HVs, comprises a bpNLS, which depends on the presence of a lysine at position 428 (or 432 in VZV).

3.3.1 The choice of cargo protein can alter the outcome of the assay

To confidently identify functional NLSs in VP1-2 a good screening system had to be chosen. The standard way to confirm NLS activity in a particular sequence is to monitor the localisation of fusion proteins of the test sequence and an easily tractable cargo. One common choice is GFP. Although this system lends itself to live cell imaging there are some notable pitfalls. Its relatively small size of 27kDa allows GFP to enter the nucleus without the need of a NLS (Figure 9B, panel I). Another choice was the bacterial protein β-gal, which is large (130kDa) enough to be excluded from nuclei (Figure 9C, panel IV) and does not contain any putative NLS or NESs making it a good candidate cargo. The latter point would make it an ideal cargo to find even weak NLSs as, once nuclear, it should theoretically remain there unless nuclear integrity is disrupted. Initial comparisons of HSV-1 NLS regions fused to both GFP and β-gal showed some disparities in trafficking in the GFP/β-gal fusion proteins with R4 and R2. Firstly, R2 was able to pull only GFP into sub-nuclear compartments which might be nucleoli (Figure 9, panel III). Secondly, the nuclear transport activity of R4 would likely not have been noticed using the GFP system (Figure 9, compare panels I and II). What are the reasons for the differences observed for R4 or R2 targeting activity? One of the following two theories might explain the GFP/β-gal sub-nuclear compartmentalisation observed in the context of the R2 fusions. (i) GFP contains additional determinants that allow it to stick to components of or enter nucleoli which β-gal does not. (ii) If one assumes that R2 contains additional features which target the protein to the nucleolus, then β-gal could simply be too large a protein to access this protein dense structure. The discrepancy around NLS activity of R4 might touch upon similar problems. GFP is small enough to enter the nucleus but it may by definition then also freely re-enter the cytoplasm. BpNLSs are formed by two basic sequences and thus R4, containing only one of the basic patches, might simply be less efficient at nucleocytoplasmic shuttling than R2 due to weaker binding to importin-α. This in turn would be exacerbated by the tendency of GFP to exit the nucleus again. In comparison, β-gal could

- 103 -

Identifying NLSs of VP1-2 homologues not exit the nucleus and thus might accumulate in the nucleus ‘more effectively’ as part of a R4 fusion protein. Thus the β-gal system might allow for more accurate identification of even weaker test NLSs and was the preferred choice for the purpose of the work presented herein. In addition to the β-gal system, NT6 was employed to more faithfully mimic the spatial and structural constraints of VP1-2. Effectively, NT6 was the more stringent system to assess NLS function since the structure mimics the native NLS position of the homologues more closely than β-gal.

3.3.2 All VP1-2 homologues contain a NLS

Using the sequence alignments and NLS prediction software as a guide I was able to show in two systems that the VP1-2 homologues from human representatives from all HV sub-families feature a functional NLS in the N-terminal linker immediately downstream of the USP. The functional determinants can be narrowed down to the first positively charged region comprising two basic patches (see R1 and R4’ in the alignment in Figure 8 and results in Figure 10 and Figure 11) for α, β and γ1 sub-family members. HHV-8, and at least two other γ2 members (Figure 12), appeared to be the exception to the rule. The HHV-8 NLS lies somewhat downstream of the first basic region and exhibited different organisation. Although some minor conservation of the basic residues remained after some manual manipulation of the alignments only one basic patch was present which might represent an adaptation to the viral life style. As part of this thesis I presented two systems which I used to study nuclear import of VP1-2 homologues; the first using β-gal and the second using HSV-1 VP1-2 (NT6). With the β-gal system it was virtually impossible to test relative strength of test NLSs since it cannot exit the nucleus, it is not live tractable and cannot be bleached precluding their use in assays of fluorescence recovery. Using NT6, however, I was able to make some limited conclusions regarding NLS strength and utility. The data presented here indicate that all NLSs with a leader lysine similarly efficiently translocated a large cargo protein (NT6) to the nucleus. Also signals with two basic patches and a leader lysine were tolerant of mutations of key residues. As such the K428R substitution and the VZV.R1, which did not perfectly match the full bpNLS consensus, were still active (Figure 15). Since NLS activity correlates with importin-NLS binding strength (Hodel et al. 2006) it could be possible that the arginine at the first position (e.g. VZV.R1) cannot make a similarly strong connection to one of the importin-α binding grooves which could explain why in R4 the K428R abolished nuclear transport. Other exceptions from the consensus have been reported. For example, the bpNLS of CBP80 comprises three arginines in the first patch and still make functional contacts with the minor NLS binding groove of importin-α (Dias et al. 2009). This negative effect was likely exacerbated by the K428A mutation which lost complete activity and thus presumably indicates loss of interaction with the relevant import factor. Similarly, when R5 was removed from NT6, nuclear import efficiency dropped likely indicating

- 104 -

Identifying NLSs of VP1-2 homologues weaker binding to importin. This could also explain why in PRV both basic patches were required for full activity of the NLS (Möhl et al. 2009). Taken together, the NLS-importin interaction probably requires both patches, although R4, but not R5, might be able to weakly interact with the nuclear transport machinery in isolation. To test this proposition a combination of biochemical approaches may be useful. Wild type and mutant peptides can be coupled to a solid support (e.g. beads) via a tag and this can be used in vitro as a bait to bind a particular importin. The complex could be subjected to increasingly stringent washes which should give an indication how strong their interaction is. Using the Biacore system, affinity of different NLS peptides to a particular importin can then be measured.

3.3.3 What is the nature of the NLS?

I initially worked on the assumption that HSV-1 comprised a mpNLS. Due to the striking conservation in R4 I anticipated that all other α-HVs would feature similar NLS properties. Thus far six classes of classical NLSs (Table 1 in 1.3.2) have been identified (Kosugi, Hasebe, Matsumura, et al. 2009; Dingwall & Laskey 1991). The α-HV R4s and the first basic patch of most β-HV and γ1-HV R4’ regions would, according to this classification, be defined as class 1 mpNLS with a leader lysine followed by three basic residues, at least one of which has to be an arginine. At least for HSV-1 and PRV R4 NLS function was experimentally confirmed (Abaitua & O’Hare 2008; Möhl et al. 2009). There are notable exceptions to this class 1 rule; at least two strains of VZV comprise five consecutive arginines; bovine Herpesvirus 1 (BHV-1) features only four arginines, preceded by KR motif some four residues upstream; MuHV2 only contains three consecutive basic residues with the first being an arginine and an additional lysine two residues upstream. Unexpectedly, the analysis of NLS function of both the VZV and HSV-1 homologues gave rise to some puzzling results, which did not fit this initial hypothesis. Firstly, VZV.R4 lacked nuclear transport activity which was recovered by extending the test region to R1. Secondly, the R432K (VZV numbering) substitution resulted in a gain-of-function of VZV.R4 in the nuclear transport assay. Thirdly, the K428R substitution in HSV-1 abolished nuclear transport of β-gal only in the context of R4 but not R1. And fourthly, in NT6, the K428R mutation showed similar efficiency as VZV, and, finally the K428A mutation completely abolished nuclear transport of NT6. This prompted me to revisit the sequence analysis and to consider the possibility that the α-HVs might feature a bpNLS. A closer look at the R1s of the β and γ1 sub-family revealed features reminiscent of a classical bpNLS with spacers between 8-14 residues. The presence of a bpNLS was recently confirmed for at least HCMV (Brock et al. 2013), although it has to be noted that both basic patches in HCMV clearly adhere to the mpNLS consensus. While the HCMV report (Brock et al. 2013) formally excluded the

- 105 -

Identifying NLSs of VP1-2 homologues presence of a mpNLS, such activity might have gone unnoticed since the authors used GFP-NLS fusion proteins, which accumulated in the nucleus even in the absence of an NLS similarly to the results presented in Figure 9B (panel I). The α-HV spacer length coincided with the length of sequence that would be necessary to bridge the distance between the two NLS binding grooves of importin-α (Kosugi, Hasebe, Matsumura, et al. 2009). In contrast to this, the bpNLS in the α-HVs showed considerable deviation from the consensus (Figure 13D) although it would be impossible, without further experimental information, to say which part of R5 was contributing to NLS function. Strikingly, at least 15 different sequences of different HSV-1 strains/isolates showed perfect conservation of R5 (see 10: Appendix 2) indicating that potentially all of R5 would be required. Theoretically though, the NLS spacer could be as short as 16 residues in HSV-1 although it likely counts at least 25 residues if one considers which basic residues are conserved between the very similar HSV-1 and HSV-2 homologues (Figure 8A). The most extreme version of NLS spacer found in the α-HVs would feature 40 residues for VZV (see Figure 8 and Figure 13D). However, for each of the bpNLS combinations at least one residue did not match the consensus and even dramatic changes of charge were observed (Figure 13D, blue letters, e.g. R>T or K>S). This may not be a problem per se, since extremely long spacers with considerable sequence flexibility have been characterised previously (Mclane et al. 2008; Moore et al. 1998; Lange et al. 2010). The reports suggest that spacer residues would form a NLS scaffold and some of the amino acids would make contact with importin-α outside the NLS binding grooves. This would make some spacer residues equally important NLS constituents, although it was shown that spacers are more accommodating to mutations than the key basic residues (Lange et al. 2010; Kosugi, Hasebe, Matsumura, et al. 2009). Despite the considerable diversity of the α-HV spacer the R4-proximal sequence is highly conserved. This might indicate that this region contributes important functions to this region which may relate to interactions with importin-α sub-types. The α-HV spacer comprises features also highlighted by Kosugi et al. (2009). Of note would be an enrichment of prolines (Figure 8, grey background) at either end of the spacer and acidic residues (Figure 8, blue background with white letters) in the central region (part of the PEST sequence). Both can be seen in VZV and to a lesser extent in the other α-HVs. PEST residues would presumably contribute to disorder and thus allow presentation of the NLS. The peptide screen presented by Kosugi et al. (2009) did not account for potential phosphorylation events. Since acidic residues can mimic the phosphorylated state of serine or threonine the acidics identified in the central region of bpNLS spacers could also represent phosphomimetics. The R1 spacer of all HVs comprises many potentially phosphorylatable residues (Figure 8, yellow background) which, upon phosphorylation, could resemble the properties of acidic residues and augment NLS activity. Phosphorylation has previously been shown to allow regulation of NLS activity. For example, the strong mpNLS of SV40 lTag was

- 106 -

Identifying NLSs of VP1-2 homologues confirmed to be activated by casein kinase (CK) 2 and CDK phosphorylation (Hübner et al. 1997; Rihs et al. 1991; Xiao et al. 1997). A putative CK2 site, namely SSVEDL (or slightly different), is conserved in the α-HV spacer and between 7-12 additional putative phosphorylation sites for various kinases (including Ataxia telangiectasia mutated (ATM), CK2 and Aurora) are predicted in HSV-1.R1 using several algorithms (Wong et al. 2007; Xue et al. 2011; Xue et al. 2008; Blom et al. 1999). One could envisage that the presence of multiple phosphorylation sites could be a means to augment or to allow regulation of NLS activity depending on the cell type and/or the phase of infection (e.g. entry versus assembly/egress). To gain a more detailed understanding of the functional determinants of the HV NLS it would be important to include more VP1-2 sequences into the in silico analysis. Differences between very closely related virus species or even strains of the same virus (although no differences were found between 15 HSV-1 strains as shown in appendix 2) might give an indication of which parts are actually contributing to function. Any notable conservation or sequence deviations can be investigated with the herein established methods. Conserved residues could be studied by alanine scanning, where one or more residues are substituted with alanine. This would be useful to assess which features of R1 contribute to function in nuclear transport assays and which mutations prevent complementation of a VP1 2 null virus. Other mutational studies (phosphomimetics, S/T>E substitutions) could yield information on regulation of this region and whether this affects the ability of HSV-1 to infect target cells.

3.3.4 An improved system for screening NLSs

There are many options to screen NLSs but which one is the most suitable? Besides identification of NLSs it might also be required to judge their strength. At the very least a nuclear localisation assay should consider some basic aspects. (i) The NLS needs to be exposed so fully folded cargos where the tagged end is not buried in the structure (ii) Mimicking the position of the NLS in the native protein is desirable (i.e. between two domains or at the termini) (iii) The cargo should be mostly, if not exclusively, cytoplasmic in the absence of a NLS (iv) The cargo should be reasonably inert, that is, not form cytoplasmic or nuclear complexes that could anchor the protein in either compartment (v) The cargo should be easily tractable (e.g. fluorescent or via good antibody) Although my study revealed NLSs in all VP1-2 homologues the data could be more comprehensive in respect to efficiency. Some minor modifications could even allow live screening of

- 107 -

Identifying NLSs of VP1-2 homologues nuclear transport inhibitors. In retrospect GFP would have been a useful due to its life tractability. In order for GFP to become a better option for this kind of assay, its size could be increased beyond the assumed limit of free entry/exit from the nuclear compartment (>40-60kDa). This could be done by fusing several GFPs in tandem, or fusing GFP to NT6. GFP-NT6 would be the most ‘physiological’ system since the test NLS would be presented either in context of VP1-2 structure. Alternatively, the test NLS could be inserted between GFP and β-gal to mimic the position between two folded domains (USP and body of VP1-2). Such a GFP based system would allow me to better analyse the strength and efficiency of the different NLSs by enabling another technique, FRAP (fluorescence recovery after photobleaching). For this, nuclear fluorescence (GFP fusion) is microscopically bleached and the time of recovery (that is, import of new protein) measured. Ultimately, a HV study should not only aim to understand replication but also to improve current anti-viral treatments since resistance to current antivirals is an increasing problem in immunocompromised patients. Knowing how all HV capsid dock at the nuclear pore should enable the development of chemical agents that specifically inhibit nuclear transport of HV capsids by blocking (binding) the exposed NLS in VP1-2 or the importin-VP1-2 complex. Using chimeric NT6 as a model, and with a few minor modifications to the construct, this could be done in culture for any HV. Ideally an inhibitor of nuclear transport should only be added briefly to avoid cellular toxicity. When NT6 is expressed in cells and the inhibitor were added after 24 h of expression, most of NT6 would already be nuclear. Thus it would be best to design a cargo that actively shuttles between the nucleus and the cytoplasm. Fusing a CRM1-dependent NES to the C-terminus of NT6 would enable export of NT6 from the nucleus (and cytoplasmic accumulation) in the presence of a nuclear transport inhibitor with only very short treatment times. Combined with good controls for nuclear transport this could guide the development of nuclear transport inhibitors with anti-herpetic activity. In addition to screening potential inhibitors of HV infection, in the described system (NT6 shuttling vector including a NES) it would provide a tool to empirically determine the nuclear transport factors that could bind the NLS. Rather than knocking-down importins, dominant negative importin-α sub-types could be expressed in cells also expressing NT6. Dominant negative importin-α lacks the importin-β binding domain (IBB) which is essential for the nuclear transport process and cargo release (Harreman et al. 2003). This would cause cytoplasmic accumulation of NT6 (or a chimera) in the presence of a particular dominant negative importin-α. This system also lends itself to pulldown studies since the interaction of a cargo to the dominant negative importin-α is stronger than to wild-type importin-α (Fanara et al. 2000) and cannot be disrupted since that would require nuclear import, importin-β binding and combined activity of the IBB and nucleoporin (Harreman et al. 2003; Gilchrist et al. 2002). That means the interaction would be stable during the pulldown procedure rather than

- 108 -

Identifying NLSs of VP1-2 homologues transient and the fraction of NT6 bound to dominant negative importin-α would essentially be higher than the fraction that is actively transported to the nucleus at any given time.

- 109 -

Analysis of NLS function during HSV-1 replication

4 Analysis of NLS function during HSV-1 replication 4.1 Studying R1 function during HSV-1 replication

My hypothesis was that all VP1-2 homologues contain a NLS at the N-terminus, which enables incoming capsids to dock at the nuclear pore and thus is a pre-requisite for infection. Since it was already established that the HSV-1 R1 region was a functional NLS and absolutely required for infectivity (Abaitua & O’Hare 2008) it made sense to initially test the homologues for the presence of a functional NLS and then proceed to analysing their function during HSV-1 replication. This way I could concentrate on regions comprising NLS function rather than screen several untested candidate regions per homologue in parallel. When the first basic patch in VP1-2 is deleted (VP1-2ΔNLS virus) the resulting virus virtually does not replicate and has to be grown in complementing cells. Complemented ΔNLS virus can enter susceptible non-complementing cells from where it subsequently cannot spread further. Ultrastructural studies revealed that ΔNLS virion assembly in non-complementing cells is comparable to wild type virus. The resulting ΔNLS virions can fuse with uninfected cells and capsids migrate along microtubules towards the nucleus. In contrast to wild type capsids, which can be seen at the nuclear pore, ΔNLS capsids do not seem to progress past the MTOC, which implies that the sole function of the NLS region is capsid docking at the nucleus (Abaitua et al. 2012). Thus it should be possible to test if NLS sequences can rescue this defect by supplying the necessary function. The only way to obtain infectious ΔNLS stocks is to grow it in VP1-2 complementing cells (e.g. RSCUL36 or HS-30). Rarely, revertants contaminate the purified ΔNLS stocks, which arise through recombination with the cellular copy of VP1-2. It should be possible to exploit this phenomenon for the purpose of studying the ability of test (NLS) sequences to rescue replication of the ΔNLS virus. The assay described in Figure 16 was based on the defect of non-infectious VP1-2.ΔNLS virus. From purified stocks the viral genomic DNA carrying the VP1-2ΔNLS mutation can be purified by removing the viral envelope and digesting the tegument and capsid components with proteinase K. If pure ΔNLS viral genomic DNA were transfected into non-complementing cells replication should progress normally. Any assembled virions would then carry VP1-2ΔNLS rendering them non-infectious. Theoretically, if a vector, containing a test NLS flanked by homologous sequence, were co-transfected with the viral genomic DNA, recombination could occur between the homologous regions (Figure 17A). In turn, VP1-2 (containing the test NLS), expressed from the recombinant viral DNA, would decorate viral capsids, but only if a test region can fulfil the same function as HSV-1.R1 would the virus spread through the monolayer of cells and give rise to visible plaques. Most importantly, the NT6 vector used

- 110 -

Analysis of NLS function during HSV-1 replication for recombination cannot complement the VP1-2ΔNLS defect in trans since it lacks the essential C-terminal domain which is absolutely crucial for capsid binding (and thus VP1-2 function). Although in theory this strategy should work, there are a few things to consider. The region I exchanged counts over 70 residues, which, although not conserved across the sub-families, might contain additional functions. If this approach were to fail, it would be possible to shorten the insert of foreign sequence in order to keep as much HSV-1 sequence a possible (i.e. on change basic region or part of it). Additionally, the GC content of the inserts can be up to 20% lower (VZV.R1) compared to the HSV-1 sequence which might impact recombination or the stability of this DNA region. In that case, the insert could be codon optimised to more closely match the HSV-1 sequence. Several factors that are difficult to judge could also affect steps after potential recombination. For example, the insertion of different R1s could affect folding of full-length VP1-2 or modulate its stability which could reduce or abolish viral replication. This may be less likely since NT6 expressed similarly well in transfected cells implying stability and/or folding are not an issue.

Figure 16 Schematic representation of the ΔNLS rescue assay. (A) The rescue assay was based upon recombination of the non-replicating VP1-2.ΔNLS virus with NT6s containing foreign NLSs. This could lead to rescue of the ΔNLS phenotype if the motif allows docking of capsids at the NPC. (B) Viral genomic DNA isolated from extracellular virus stocks and NT6 plasmid DNA were co-transfected into non-complementing RSC cells.

- 111 -

Analysis of NLS function during HSV-1 replication

Viral replication occured from the viral genomic DNA, but only if recombination, which inserts a functional sequence, took place, can infectious virus be produced. NT6 protein itself lacks the highly conserved, essential C-terminal domain. This means that trans-complementation cannot be the reason for release of infectious virus. If virus is recovered, it will be serially purified under 0.4% agarose overlays, sequence confirmed and characterised.

4.2 Function of homologous R1s during HSV-1 replication

To validate the rescue assay format viral genomic ΔNLS DNA was either transfected alone or in combination with NT6 wild-type or NT6.ΔNLS. After three days of infection, only the plate where NT6 wild-type had been co-transfected with viral genomic ΔNLS DNA showed CPE (Figure 17A, compare plates 1-3). When virus preparations from these control plates were titrated, only samples where NT6 wild-type had been co-transfected contained detectable virus (Figure 18B/C). If my hypothesis were true, then the homologous R1s should allow plaque formation. In order to test this chimeric NT6 vectors were co-transfected with viral genomic ΔNLS DNA. Only if recombination into the viral DNA occurred (see Figure 17A) would cytopathic effect be observed. As for the controls (Figure 18A, plates 1-3), viral genomic ΔNLS DNA was transfected in combination with the chimeric NT6s (plates 4-7). Cells were then left for three days to allow viral replication to occur. After fixation of the co-transfected cells plaques were observed for VZV.R1, CMV.R1 and EBV.R1 (Figure 17A6). Using exactly the same amount of viral genomic ΔNLS and plasmid DNA for each condition allowed a direct comparison of rescue efficiency. Rescue was observed for three out of five chimeric NT6s (including the two different NT6s for HHV-8.R1 and R8) and most plaques were obtained with NT6.VZV.R1, fewer with EBV.R1 and CMV.R1. In contrast to that, there were no plaques observed when the HHV-8 chimeras were co-transfected with viral DNA (Figure 17A, plate 7). The result of plaque formation was confirmed in the virus yield experiment. Cells transfected with the VZV chimera yielded approximately 37-fold fewer infectious virions compared to wild type NT6. However, in comparison to the other NT6 chimeras VZV gave rise to 6 fold more infectious virus than the EBV chimera and 15-fold more than the CMV chimera (Figure 17C). Despite the drop in titre compared to wild type, virus could be purified from these preparations, which was confirmed by sequencing the insert. This leads to the conclusion that those regions that comprise a bpNLS were able to rescue the HSV-1 ΔNLS mutant while the mpNLS from the HHV-8 chimera was not functional in the context of HSV-1 VP1-2.

- 112 -

Analysis of NLS function during HSV-1 replication

Figure 17 Rescue assay quantification and characterisation of VP1-2 integrity. (A) ΔNLS DNA was co-transfected with NT6 plasmids into non-complementing RSC cells and monolayers fixed and stained with crystal violet at 3 dpi. DNA alone and in combination with NTΔNLS exhibited no plaque formation, while NT6 wild- type showed complete cytopathic effect (CPE). Plaque formation was observed for, in descending order of efficiency, VZV.R1, EBV.R1 and CMV.R1 in contrast to the HHV-8 sequences. (B and C) Cells from one well of equivalent co-transfection experiments were scraped into their respective media, and virions liberated by three to six freeze-thaw cycles (i.e. pooled virus). Cleared lysates were titrated on non-complementing RSC or Vero cells. Plaque numbers were quantified. (D and E) NT6 plasmids were co-transfected with HA-tagged UL37 expression vector and NT6 immunoprecipitated using anti-V5 antibody. IP eluates were compared for the presence of HA-UL37 by Western Blot using anti-HA antibody.

- 113 -

Analysis of NLS function during HSV-1 replication

4.3 Functional determinants of HSV-1 R1 required for replication

To establish which parts of HSV-1.R1 were important for viral replication, the rescue assay was repeated with the mutated HSV-1 NT6s (Figure 17B). Initially, the for localisation more dominant region R4 was investigated. Not surprisingly, the K428A mutation, already found to abolish NLS activity in R1/4.β-gal and NT6, did not support replication, whereas the conservative K428R substitution gave rise to viral titres similar to that of VZV.R1 (Figure 17B). This points to a role of NLS strength during replication as by substituting the first lysine with an arginine both nuclear localisation and viral fitness appeared impaired. Exchanging the basic residues in R4 for a strong cellular NLS allowed for recovery of virus albeit at a 100 fold lower titre than wild-type. This underscores the complexity of capsid nuclear transport. The K428R result might imply that the reduction in titre was caused by the reduction in NLS activity. This could then be extended to VZV.R1 since the obtained titres were very similar. However, while this conclusion might be true for the reasonably well conserved α-HVs it cannot be extended to the R1s from the β- and γ1-HVs. Pure strength of the NLS only augmented viral replicationif integrity of R1 was maintained since NT6 with the MyoNLS insert was almost exclusively detected in nuclei of transfected (Figure 15B) cells but gave rise to lower titres than K428R or VZV.R1 (Figure 17B). Despite the loss in recovery, recombinant viruses were purified for both K428R and MyoNLS mutants. To establish the role of R5 the assay was repeated with a NT6ΔR5 mutant. Even though R4 was intact in this construct it did not complement the replication defect of the ΔNLS virus. This was surprising since in comparison to NT6.K428A NT6ΔR5 retained some NLS activity at least in HuH7 and COS cells (B/C). Intriguingly, in HEK293 cells NLS activity was similar for both K428A and ΔR5 and only marginally above ΔNLS control. This might indicate that using a different cell type might change this result although this seems unlikely.These data add to the aforementioned conclusion that NLS strength only augmented capsid nuclear entry if the whole R1 region was intact.

4.4 Analysing pUL37 interaction with chimeric VP1-2

Since the strong NLS sequence HHV-8.R8 was unable to rescue virus growth it was important to exclude any deleterious effects the presence of foreign sequence could have on functions of adjacent parts of VP1-2 (see Figure 3 in chapter 1.2.1). Possible problems could arise in respect to USP function (cleaving ubiquitin from target proteins including VP1-2 itself), VP16 binding, pUL37 binding and folding of the middle segment. USP function was likely not impaired since NT6 expressed well and to levels comparable to wild type NT6 (Figure 17D/E). Any functional impairment of the USP would have destabilised VP1-2 (Bolstad et al. 2011). VP16 is at least partially recruited to capsids by VP1-2 but since there is approximately 5-

- 114 -

Analysis of NLS function during HSV-1 replication fold more VP16 than VP1-2 in virions disruption of this binary interaction likely would not cause complete loss of infectivity. In fact, a binding deficient VP16 mutant virus was found to be viable, albeit exhibiting a modest defect (Svobodova et al. 2012; Newcomb et al. 2012). Folding of the middle segment was probably also not affected since the exchanged region outside the NLS shows little conservation till approximately 100 residues downstream of R1. Additionally, unfolded proteins are more prone to degradation and aggregation but the levels of all NT6 isoforms were very similar (Figure 17D/E). The most likely culprit would thus be pUL37, which binds VP1-2 at a site mapped to a region encompassing R5 in PRV (Fuchs et al. 2004). The pUL37-VP1-2 interaction is essential to HSV-1 replication and pUL37 was found to contribute multiple functions including improving microtubule transport (Pasdeloup et al. 2012; Krautwald et al. 2009) and facilitating secondary envelopment (Jambunathan et al. 2014). To test whether R1 exchange could have impaired this interaction, NT6 isoforms and HA-UL37 were co-transfected into HuH7 cells and NT6 was immunoprecipitated using anti-V5 antibody (Figure 17D/E). In all cases it was found that HA-UL37 was detected in eluates from protein A coated beads. This was entirely dependent on the presence of NT6 (Figure 17D/E, lane 2 and 13) or anti-V5 antibody (lanes 5, 10 and 16) but not on the NLS region. There were similar amounts of NT6 and HA-UL37 precipitated for all constructs tested, including both HHV-8 chimeras (see lanes 20/21). This indicates that adjacent functions were likely not affected by the insertion of R1s, although there might be slight differences in function of both pUL37 and VP1-2 in RSC cells, where the rescue assay was performed, compared to HuH7, in which the pulldown was performed.

- 115 -

Analysis of NLS function during HSV-1 replication

4.5 Replication characteristics of recombinant viruses 4.5.1 Analysis of virus production

To study the role of the different NLS regions during viral replication the rescue assay has limited scope. Although the assay yielded valuable information on viability of the different NLSs, it cannot really be used to make any conclusions about fitness of a particular recombinant virus. There are just too many factors that could have influenced the efficiency of recombination. Nonetheless, the results of the rescue assay indicate that recombinant viruses may be recovered from transfected cells. To obtain recombinant virus was a simple procedure, which involved serial passage of individual virus plaques to reach 100% purity similar to picking a bacterial colony for cloning. Thus to obtain recombinant viruses, the rescue assay was repeated with the conditions that previously gave rise to plaques. This time the cellular monolayer was scraped into the medium and any cell bound virus released by freeze-thawing. This mixture was subsequently used as inoculum to infect non- complementing cells with a limiting dilution. Through the use of agarose (in the medium) it was possible to pick individual plaques and repeat this procedure at least once more. After two passages a stock was produced and the region around the NLS sequenced. These viral stocks were then used to test how the foreign NLSs compare functionally during HSV-1 viral replication. For this virus yields from single and multi-step infections were determined, which give an indication of virus production and infectivity.

4.5.2 A cell-type specific defect in replication linked to the VP1-2 NLS Multi-step growth characteristics

A multi-step growth curve measures how well a particular virus spreads through a monolayer of cells over time by evaluating the production of progeny virions. It yields information on a combination of virion production and infectivity, since only one in a thousand cells were initially infected. This means the virus then has to go through several replication cycles including virion production and re- infection. This is eventually measured as virus yield after 3 dpi, the time when a wild type virus would usually have caused 100% cytopathic effect. For the multi-step yield experiments summarised in Figure 18A-D, four different cell lines were infected at a MOI of 0.001 and harvested every day for three days. In RPE-1 cells (Figure 18D), a cell line that should mimic naturally infected cells of the eye more closely, yield was not affected for any of the recombinant viruses at any time point, except for the final yield for the MyoNLS recombinant virus. Replication of the MyoNLS recombinant virus in RPE-1 cells was modestly impaired compared to

- 116 -

Analysis of NLS function during HSV-1 replication wild-type virus (around 5-10 fold lower titre). These data indicate that the factors required for capsid transport by the different NLSs were likely present in these cells. In RSC cells recombinant virus yield after 24h was up to 100-fold reduced for EBV.R1, CMV.R1 and VZV.R1 compared to ΔNLS revertant virus (wild type) and approximately 50 fold lower than that of KOS wild-type virus. By 48h the titres of EBV.R1 reached levels around 2-5 fold lower than KOS wild- type whereas the difference compared to ΔNLS revertant was reduced to around 10-20 fold. VZV.R1 and CMV.R1 were up to 10 fold and 50 fold lower than KOS wild-type and ΔNLS.Rev, respectively. However, after three days in culture titres obtained from all viruses reached similar levels as the whole mono layer of cells became infected and started producing virus particles. The reduction in early yield was not completely rescued by the complementing cell line RSCUL36 (Figure 18B) and thus might indicate that the EBV, CMV and VZV R1s caused an early defect, quite possibly during entry and/or a defect during replication generally. In Vero cells (Figure 18C), a Monkey kidney cell line commonly used to passage and characterise HSV-1 growth parameters, virus yield after three days dropped over three orders of magnitude for VZV.R1 (light brown line). CMV.R1 (darker brown line) yielded slightly more virus yet still gave rise to significantly lower amount of infectious virus particles (approximately 500 fold reduction). EBV.R1 (green line) exhibited an intermediate phenotype (15 fold reduced) while the MyoNLS appeared to not significantly affect the ability of HSV-1 to replicatate. This defect was already noticeable at 24hpi (between 10 and 100 fold reduction) and got sequentially larger, which may be a combination of reduced infectivity and slower replication that in the case of a multi-step growth curve cause a knock- on effect for each subsequent infection (i.e. it takes virions longer to initiate an infection that again yields fewer infectious particles which then go on to repeat this cycle).

- 117 -

Analysis of NLS function during HSV-1 replication

Figure 18 Multi- and single-step yield of recombinant HSV-1 from four different cell lines. (A-D) Yield from RSC, RSCUL36, Vero and RPE-1 cells infected for up to three days with low MOI (0.001) was quantified on complementing cells after the indicated time points. Cells and supernatant were harvested simultaneously by scraping cells and freeze-thawing the mixture 3-6x. (E) Monolayers of the same cell types were infected with MOI of 10 and virus particles harvested 24h later (as in A-D). Infectious virus yield was determined on complementing RSCUL36 cells. Single-step growth characteristics

The significant reduction of virus titres observed in the multi-step growth curve revealed a defect at some stage of replication but this information was not fully informative. This could have been

- 118 -

Analysis of NLS function during HSV-1 replication caused by lower virus production in infected cells, by a defect to establish infection or a defect during assembly and egress. To test if lower virus production could have caused both the early reduction in multi-step yield observed in RSC cells and the overt defect observed in Vero cells I infected the same cell lines with a MOI of 10 (Figure 18E). Since this ensures that almost all cells are infected roughly at the same time this allows to check how much virus a cell produces on average during one round of infection (here 24h). In RPE-1 cells VZV.R1 and CMV.R1 (Figure 18E, light and dark brown bars, respectively) showed a very modest reduction in infectious virion production (up to 5 fold) whereas EBV.R1 (green bar) gave rise to comparable infectious virus as KOS wild-type. The recombinant virus carrying the MyoNLS insert (orange bar) appeared to replicate less well yielding around 20 fold less infectious virus, which may have caused the modest reduction in virus titre observed in multi-step growth experiments. These results were in agreement with the multi-step growth curve. In RSC cells (Figure 18E) a modest reduction of up to 20 fold could be observed for CMV.R1 (dark brown) whereas VZV.R1 (light brown) was only 5 fold reduced. EBV.R1 (green) and MyoNLS (orange) exhibited only minor differences compared to both wild-type viruses. The modest reduction of progeny virion production for VZV.R1 and CMV.R1 could explain the initial slow progression in the multi-step infection. Intriguingly, in complementing RSCUL36 cells the modestly impaired virus production was not rescued beyond the defect in RSC cells. This could indicate that the effect of the chimeric VP1-2 is dominant and would require more wild-type VP1-2 to effectively compete with mutant VP1-2 for capsid binding. Compared to the other cell lines, Vero cells show quite a large defect in progeny virion production for all chimeras although the virus carrying the MyoNLS replicated reasonably well (approx.. 10-fold lower yield). A greater than 1000-fold decrement of infectious virions that can be recovered after 24 hpi, could indicate a major defect during assembly. However, the lower yield could also just reflect a prolonged eclipse phase, that is, extremely slow entry and replication kinetics, which would reduce the amount of virus released compared to wild type virus (at most time points).

4.5.3 Analysis of plaque size

Conventionally, plaque assays are used to determine the titre of a viral stock. However, there is additional information that can be extracted from this assay. During a plaque assay a monolayer of cells is inoculated. After initial attachment and envelope fusion occurred, anti-HSV neutralising antibodies were added to the medium which prevent extracellular spread of infection, that is, virus could only spread laterally from cell-to-cell. The size of a plaque, a measure combining virion production and infectivity similar to the multi-step growth curve, can then be measured and compared

- 119 -

Analysis of NLS function during HSV-1 replication to wild type virus. To see whether the defect in virus production also had an effect on cell to cell spread the plaque size of several chimeric viruses was compared. For this assay, four cell lines (HaCaT, Vero, RSC and RSCUL36 cells) were infected with a low MOI. After three days plaques were detected by immunoperoxidase staining, a technique also used for immunoblots and more sensitive than counter-staining cells with crystal violet (commonly used for plaque assay). In Vero cells reasonably large plaques are formed by wild-type viruses which showed notable cytopathic effect. The different chimeric VP1-2s clearly led to a reduction of plaque size but most significantly for VP1-2.VZV and CMV (Figure 19B). For these recombinant viruses only up to 10-20 cells were infected which could only be detected using immunoperoxidase staining. In contrast to this the K428R mutation gave rise to easily visible plaques but nonetheless much smaller than wild type virus. This indicated that in Vero cells NLS strength in the context HSV-1.R1 was a determining factor. However, the defect of the VZV and CMV chimeras could not be explained by this since the VZV NLS harboured an equally potent and CMV.R1 a much stronger NLS than K428R. The VP1-2.EBV and VP1- 2.MyoNLS mutants also presented with small but visible plaques which were bigger than those of VZV, for example. It has to be noted that the overall plaque size difference between the different chimeras is small compared to the difference in relation to wild-type virus. To make this analysis more relevant to HSV-1 infection, plaque size was also compared on HaCaT keratinocytes, which should more closely resemble another physiological host cell type (Figure 19A; bottom two images and Figure 19B; middle graph). Because these cells produce so much extracellular virus it is difficult to quantify exact size of a plaque for wild-type virus. On these cells extracellular virus production is large enough to allow virus spread through the medium, forming so called comet tails (Figure 19A; red arrows), despite the presence of neutralising serum (here 1-2% human serum). Thus plaque size could only be estimated for wild type virus (Figure 19A, see red circles) but cannot be used as a real measure. However, it is clearly visible that size is larger than that of the biggest chimeric virus (EBV, Figure 19A/B). All chimeras tested showed a defect in cell-to-cell spread reflected in more than 60% reduction in plaque size (>60% for EBV, >75% for VZV, >85% for CMV and >90% for MyoNLS), suggesting the virus production and/or infectivity could be impaired in HaCaT cells. Unexpectedly, the K428R mutation showed only a modest reduction of plaque size of around 30% indicating that NLS strength has some limited effect on infectivity and/or virus production in the context of the HSV-1 motif only. As mentioned before comet tails arise even in the presence of neutralising serum. These were differentially affected in the different chimeras (see the middle graph in Figure 19B, red plus or minus). The results indicate a reduction in extracellular virus release. Most significantly, VP1-2.Myo and VP1-

- 120 -

Analysis of NLS function during HSV-1 replication

2.CMV recombinants did not yield any detectable comet tails while the VP1-2.EBV and less so the VP1- 2.VZV chimeras showed small tails of infected cells (for EBV comet tail examples see Figure 19A). This indicates that extracellular virus production and plaque size are certainly linked and the observed defect in plaque size and comet tails formation might might have been caused by a general reduction in overall production of infectious virions and infectivity (see section 0). To check whether the reduction of plaque size was dependent on mutations within VP1-2 I wanted to compare plaque size of all chimeras on complementing RSCUL36 cells and non- complementing RSC cells (Figure 19B; right graph). The VP1-2.EBV chimera formed the largest plaques (of all mutants) which were more than 30% smaller (area) compared to wild-type virus on the non- complementing RSC cells (Figure 19A; top two images). Again wild-type virus forms large merging plaques and cytopathic effect was visibly enhanced compared to the EBV chimera (see cleared area within red circles). Expectedly, plaque size was rescued for all recombinant viruses (at least double except EBV which was 1.6x increased). The highest factor was calculated for the VP1-2.CMV chimera, but the baseline in the non-complementing cells was lower than for the other recombinants. Plaque size on the complementing RSCUL36 cells improved by a lesser factor for VP1-2.VZV and Myopodin. but these were also consistently larger compared to the non-complementing RSC cells. This clearly indicates a defect related to VP1-2. For the K428R recombinant plaque size on RSC cells was affected more than on other cell lines despite yielding similar amounts of infectious virus in single and multi step growth curves compared to wild-type. This, in combination with the reduction of plaque size of most chimeras, could indicate a specific defect only in cell to cell spread in RSC cells as virus yield was largely not affected.

- 121 -

Analysis of NLS function during HSV-1 replication

Figure 19 Plaque size comparison of the recombinant HSV-1 mutants on different cell lines. (A) Representative fields from 6 well plates showing plaques of NLS.R and of the EBV chimeric virus (the largest plaques of all the chimeric viruses). Plaque size for NLS.R was estimated due to their large size (red lines). In HaCaT cells NLS.R plaques appeared to form ‘comet tails (red arrow) due to convection currents moving virions that were insufficiently neutralised by human serum. (B) Plaque size on Vero (left), HaCaT (middle), and RSC and RSCUL36 (right) cells was measured with Quantity One software and the results illustrated using arbitrary surface units (i.e. pixel area).

4.5.4 Analysis of viral protein expression

The reason why all chimeric viruses were debilitated on Vero cells, showing reduction in production of infectious virions in single- and multi-round infections and plaque size, could be caused by slower or reduced entry into target cells. This can be tested by Western Blot to measure the time till viral protein expression can be detected when all cells in a dish are infected (high MOI infection). For this cells have to be infected as explained for the single step growth curve but it becomes vital to synchronise entry as much as possible. To achieve synchronisation, the adherence phase was carried

- 122 -

Analysis of NLS function during HSV-1 replication out on ice to prevent virus from fusing without impairing attachment. Once the temperature is elevated bound virions can fuse and infection progresses as usual. Samples were then taken periodically to monitor viral protein expression. For the purpose of this analysis I chose to compare the chimera with the smallest average plaque size to the wild type virus since it could be assumed that the defect will be strongest and analogous to that found in the other chimeric viruses. Vero (large phenotype) and RSC cells (modest phenotype) were infected with the VZV chimeric and wild type viruses at high MOI and infected cell lysates, harvested at different times, were tested for the presence of representative immediate early (ICP4) and late (VP5, pUL37, gB) proteins by Western Blot (Figure 20). Immediate early proteins could be detected as early as 3 hpi in RSC with wild type virus while only a weak band was visible for the VZV chimera at 5 hpi. The band found in VZV chimera infected RSC cells at 3hpi presumably was residual viral ICP4 from the inoculum still present in the cells. When RSC cells are infected with the most debilitated NLS virus (i.e. the NLS mutant), the cells produce non-complemented virions but the infectivity of equivalent stocks is likely drastically reduced (Abaitua et al. 2012). If the particle to pfu ratio were also increased for the VZV chimera then much more input material would enter cells per pfu compared to wild-type viruses. Since ICP4 is delivered into cells by the incoming virus the ICP4 input would be higher per pfu for a debilitated virus. A similar band could be found for pUL37, too (Figure 20B). The reason why this probably was input ICP4 (and pUL37) is that the band intensity dropped by 5 hpi only to increase again after 7 and 10 hpi. Why only ICP4 and pUL37 (but not VP5 or gB) could be found in VZV infected RSC cells is unknown. Both VP5 and gB could be more fragile proteins in the context of newly infected RSC cells. Capsids are subject to proteasomal degradation (Horan et al. 2013) after entry which would reduce the amount of detectable VP5. The amount of gB might also be more limited than that of ICP4 proteins but no evidence of this can be presented. Otherwise it still remains possible that the antibodies for ICP4 and pUL37 are just more sensitive and thus pick up even minute amounts of protein. To circumvent this problem it might be possible to check for gene expression by quantitative real-time PCR rather than actual protein. At 5 hpi the wild type virus expressed much more ICP4 than VZV. In fact, the level of ICP4 in VZV only reached comparable levels at 24 hpi indicating slower overall infection, which at this point might be attributed to slower entry kinetics. A similar phenomenon was observed with the late protein pUL37, where residual UL37 appeared to be present at 3hpi in the VZV sample, which declined to undetectable levels by 5hpi. While forwild type virus UL37 production was detectable at 5 hpi, in VZV similar levels were detected only after 10 hpi (a 5 h delay). VP5 expression which is required for capsid formation (it is the major capsid protein) started at least 10 h earlier in revertant virus and only after 24 hpi for the VZV reached levels approximately 50% of that of wild-type. The late protein gB

- 123 -

Analysis of NLS function during HSV-1 replication recapitulated the previous findings of the other proteins. It has to be noted here that in RSC cells the absolute levels of all proteins for VZV never reached that of wild-type virus. Although there appeared to be an early defect which was probably the cause of the delay in virus production in the multistep growth curve and the modest reduction in single round infection, the observed protein levels appeared to enable production of infectious progeny virus in RSC cells at least. In Vero cells the kinetics were much slower for revertant virus with only weak expression of ICP4 at 5 hpi and final levels much below that found in the RSC cells. This is found to be the case for all viral proteins tested and points to generally slower kinetics of viral infection in Vero cells. This slow-down is exacerbated in case of VZV. Here only a faint band for ICP4 was detected at 7hpi which did not increase by the same margin as found for revertant virus by 24 hpi. Expression of pUL37 was detectable at comparable times in Vero and RSC although levels were much reduced for wild type virus. For the VZV chimera the appearance of a pUL37 band was delayed by around 3 h (Figure 20, compare 7 hpi RSC and 10 hpi Vero). Analogously, VP5 expression was delayed in Vero cells for both test viruses. By 24 hpi the levels for the VZV chimerawere barely comparable to the 10 h time point of the revertant virus. Also, the levels of gB were generally lower in Vero cells for both viruses but still high enough for virion formation since high titre virus stock can readily be obtained in these cells (Figure 18E). In comparison to wild type virus gB could barely be detected in the the VZV chimera infected cells even after 24 h of infection. This could have impaired virion formation in Vero cells and thus contributed to the observed phenotype of microplaques and lower stock titre in both multi and single step growth curves. In conclusion, the reduction in ICP4 levels for the VZV was likely caused by delayed, if not impaired, entry of capsids decorated with VP1-2.VZV chimeric proteins. This in turn would lead to a significant reduction in viral proteins (including VP5 and gB) at later stages and thus lead to fewer virions being released by 24 hpi.

Figure 20 Viral protein expression of ΔNLS.R (wild type) and VZV.R1 in Vero and RSC cells. (A) Vero cells were infected at high MOI (5) and cells processed for Western Blot analysis at the indicated time points. Cell equivalent fractions were separated on SDS-PA gels and four representative proteins were detected by immunoblot. (B) As in A but experiment performed in RSC cells.

- 124 -

Analysis of NLS function during HSV-1 replication

4.6 Discussion

The data presented in this chapter suggests that NLS function was required but not sufficient for rescue of a entry defective VP1-2ΔNLS mutant. Rescue could principally be achieved when both basic patches of R1 were intact. Although the efficiency of rescue did not correlate with strength of the NLS in the chimeric VP1-2, I was able to show that viral replication is impaired if the HSV-1 NLS is weakened but was dependent on integrity of R1. Most importantly, viruses carrying the VZV.R1, HCMV.R1 and EBV.R1 could be purified and characterised revealing general debilitation which was most pronounced in Vero cells.

4.6.1 A robust assay for evaluation of NLS function in supporting virus replication

The rescue assay was designed to screen for VP1-2 R1 (or other) insertions that can rescue the defect during entry of capsids caused by the VP1-2ΔNLS deletion. It relies on recombination of foreign sequence from a transfected vector (here NT6) into the viral genome by cellular recombinases. Essentially virus can only arise after recombination since NT6 itself, when it is expressed from the plasmid, cannot complement virus replication in trans. This made it a very tight and easy to interpret screen with few false positive results except for a few circumstances when the rescue vector was sub- optimally designed. What was the reason to choose this approach over others? At least two other theoretically feasible methods exist which could have achieved similar results.

(i) Bacmid technology to introduce mutations in vitro using conventional cloning (ii) Complementing viral replication in trans by supplying full length VP1-2 chimeras

(i) Bacmid technology allows the insertion of virtually any mutation into the viral genome through a combination of PCR and recombination into the viral genome. Although very useful to introduce mutations this method involves several cloning and recombination steps that can take months to obtain a viral mutant. Passaging a large bacmid in bacteria could introduce further undesired mutations which are difficult to control for. However, this technique has the advantage in that any mutant (even non-viable ones) can theoretically be made which cannot easily be recovered with my rescue approach (Figure 17B, e.g. K428A mutation). (ii) The second option would involve using full length VP1-2 expression vectors which can be transfected into susceptible cells prior to infection. The protein would then be supplied in trans and, if the protein were fully functional, the lysate would then contain infectious virus. Despite a similarly easy set up this method has some caveats.

- 125 -

Analysis of NLS function during HSV-1 replication

Firstly, two types of virus could be obtained in this assay; trans-complemented or recombinant viruses, which would in turn complicate the isolation of true recombinant viruses which would face stiff competition from the trans-complemented input ΔNLS virus (i.e. background). Secondly, VP1-2 is not expressed to high levels from a plasmid and could lead to false negative results. Finally, this method is further complicated by the additional cloning required since no convenient restriction sites exist in VP1-2.

4.6.2 A NLS in VP1-2 is required but not sufficient for HSV-1 VP1-2ΔNLS rescue

As elaborated in section 4.1 my hypothesis was based on the assumption that the only function of the R1 region was to facilitate docking of HSV-1 capsids to the NPC by binding nuclear transport factors. Thus a foreign test NLS inserted in place of the HSV-1 NLS should allow capsids to dock at the NPC. In practise the story turned out to be more complicated. While only test regions with NLS activity rescued the ΔNLS defect, NLS function was not sufficient. As such HHV-8.R8, VZV.R1, EBV.R1 and HCMV.R1 could substitute NLS function (Figure 15), but rescue of the ΔNLS phenotype was not achieved for the HHV-8.R8 sequence (Figure 18A/C). This raised the question of why a strong NLS would not rescue replication? Either my assumptions were wrong or some additional conditions have to be met besides NLS function. To explain this conundrum one has to consider several issues which concern R1-adjacent functions including the USP, VP16 and pUL37 binding regions, fibril formation of the central body. If USP function were perturbed then it would be expected that very little NT6 could be recovered, as this domain was shown to stabilise NT6 by removing ubiquitin chains (Bolstad et al. 2011). Since all the chimeric proteins were similarly stable (Figure 17E), it seems unlikely that USP function had been affected. VP16, which is absolutely required to establish an infection, is a major constituent of HSV-1 virions but its recruitment is only partially dependent on VP1-2 (Ko et al. 2010; Kelly et al. 2009). Its recruitment to capsids plays a role during secondary envelopment (von Einem et al. 2006) which likely was not affected by insertion of a foreign R1 since the HCMV and EBV sequences worked. VP16 also likely has no function during capsid docking at the nuclear pore since it is thought to dissociate from incoming capsids (Granzow et al. 2005) and I speculate that perturbation of an entry- related function is unlikely. Alternatively the explanation could be a structural matter. In my opinion the failure to rescue can most likely be traced back to the structure of the NLS region or the type of nuclear transport factor required to capsid docking at the NPS. Firstly, the HHV-8.R8 region contains only one basic patch (mpNLS) followed by an acidic region (Figure 12C/D) and might thus pick up a different nuclear

- 126 -

Analysis of NLS function during HSV-1 replication transport factor incompatible with capsid binding the NPC. And secondly, in the context of an incoming capsid VP1-2 likely adopts a certain conformation; its C-terminus is anchored to the capsid, the central body of the protein forming a mono- or multimeric fibre presenting the N-terminus of VP1-2 to the cytoplasm, and at least the pUL37 binding region (immediately downstream of R1) being occupied during the entry processes. Although I have shown that NT6 still binds pUL37 (Figure 18E) I have not checked if the complex would still accumulate in the nucleus. This notwithstanding, the capsid bound VP1-2 complex might form a more rigid structure than in the context of isolated soluble VP1-2 and thus limit the flexibility of the interaction with a nuclear transport factor and the nuclear pore to such an extent that only certain NLS types or conformation are compatible. It has been shown that class 1 mpNLSs bind to the major groove (more N-terminal than minor groove) and bpNLSs to both grooves of importin-α to facilitate transport (Kosugi, Hasebe, Matsumura, et al. 2009; Leung et al. 2003; Conti & Kuriyan 2000). Although being part of a bpNLS, the HSV-1 R4 region conforms to the class 1 mpNLS consensus, suggesting it may bind to the major groove of importin-α when expressed alone and in the context of R1 (thus R5 binds the minor groove). This would mean importin-α would be arranged in a way that its N-terminus faces the VP16 and the C-terminus the pUL37 binding site. The HHV-8 sequence conforms to the class 4 NLS consensus, which was found to bind to the minor groove (Kosugi, Hasebe, Matsumura, et al. 2009). This might require importin-α to bind in opposite orientation which might not allow capsid-bound VP1-2 to engage a nuclear transport factor or even the NPC in a functionally relevant way. When the Myopodin NLS is inserted in place of R4, on the other hand, the virus is still able to replicate, which supports the above hypothesis, since it would not alter importin orientation. To fully solve the question of orientation and type of NLS, NT6 vectors, which will be used in the rescue assay, should be constructed which harbour:

(i) only the Myopodin NLS (no R5 and +/- the high homology immediately next to the first basic patch in R4) (ii) the Myopodin NLS in place of R5 (i.e. R4-MyoNLS combination) (iii) a modified R1 with R4 and R5 reversed (iv) only a cellular bpNLS (e.g. nucleoplasmin)

This should indicate whether HSV-1 requires a certain type of NLS (mpNLS v bpNLS) and whether the orientation of the NLS plays a significant role during entry. This could be complemented with structural studies to see how particular NLS peptides interact with importin-α and whether they bind with a preferred orientation. Although it could be concluded from these results that NLS function was not sufficient for rescue (see HHV-8) and that there was no correlation of rescue and NLS strength (compare recovery of VZV.R1 to HCMV in Figure 18C) for the chimeric VP1-2s, it was evident that NLS function was required for

- 127 -

Analysis of NLS function during HSV-1 replication rescue to occur. It is difficult to say why the weak NLS sequence of VZV.R1 yields the most plaques plaques in the rescue assay. One explanation could be the recombination itself. Since the inserted DNA sequences are approximately of equal length and do not show much conservation across homologous inserts it is unlikely that recombination was more efficient using VZV.R1 (the region outside R1 was exactly the same for each construct as well). However, the VZV.R1 sequence actually exhibits an average GC content around 20% lower than that of the HSV counterpart (while CMV and EBV R1s contain 5.5% and 3.3% fewer guanines or cytosines than the HSV sequence, respectively). A second (and more likely) reason for the discrepancy could be the overall protein conservation within R1, which is highest in VZV, and thus may provide better function in the context of HSV-1 VP1-2. As such, other cis-acting factors (VP1-2 contains C-terminal domains with essential functions) might regulate NLS activity better the more similar it looks to the wild type sequence. It also has to be considered that in the system in which nuclear localisation was assayed human or primate cells were used. The expressed repertoire of nuclear transport factors might differ in rabbit skin cells or the importin homologues might recognise NLSs differently to its human or primate homologue. The above conclusions are mostly true for the HSV-1 NLS mutants, although NLS strength impacted rescue and viral fitness in some special cases (K428R, compare Figures 16B-D and Figure 18B). This only holds true if R5 was present. As such I could confirm that deletion of R5 abolished rescue although nuclear localisation was not completely absent (Figure 18B). This corroborates data on HCMV who showed that the HSV-1 NLS could poorly complement the VP1-2ΔNLS mutant HCMV defect (Brock et al. 2013). In their study a VP1-2 construct harbouring the HSV-1 R4 region instead of the HCMV bpNLS was used. My data now explain why they were not able to obtain a chimeric virus harbouring the HSV-1 motif as part of HCMV VP1-2, since even in HSV-1 the second basic patch was absolutely essential to viral replication. As part of the HSV-1 mutational study I have shown that R4 alone is not fully functional (see Figure 18A and B, ΔR5 rescue).In agreement with the localisation data, R4 was the dominant feature for rescue since the K428A mutation completely abolished viral replication. Intriguingly, for these HSV- 1 mutants (likely not for chimeras!) rescue correlated with the presence of category I and II nuclear localisation in HEK293 cells (Figure 16D). Here NT6.K428R shows reduced nuclear localisation compared to wild type but around 40% category I/II events while for NT6ΔR5 and NT6.K428A not a single cell with predominantly nuclear localisation was found. The higher stringency (due to factors discussed in the previous chapter) in this case could have revealed that the ΔR5 NLS is more debilitated than could be anticipated from the HuH7 data. Quite often some cancerous cell lines show elevated

- 128 -

Analysis of NLS function during HSV-1 replication levels of certain nuclear transport factors (van der Watt et al. 2011) and thus a weaker NLS might be rescued while this might not be the case in cells with low expression of a particular factor. This still leaves the question why R5 is required for HSV-1 replication? This could touch upon similar aspects as mentioned with HHV-8.R8. Without R5 the mpNLS in R4 might still bind to importin-α at the same site but due to lack of R5 binding the second site on importin-α the NLS-importin complex might not adopt a conformation/orientation that is compatible with other adjacent functions (see pUL37 and VP16 comment above). These data (Figure 18B, lack of rescue with ΔR5) contrast what was found for another α-HVs, PRV (Fuchs et al. 2004). They argue that a 238 amino acid deletion within the VP1-2, which includes most of the PRV.R1 region downstream of the basic residues in R4 (sequence deleted after PKRRRP motif), still allows some replication to 50 fold lower titres. This defect was attributed to lack of UL37 binding capacity. However, I was able to co-precipitate pUL37 with NT6.ΔR5 (and all of the chimeras tested) which indicates that R5 does not contribute to pUL37 binding. In fact, their phenotype might actually relate to loss of R5 rather than pUL37 binding since, in contrast to HSV-1, in PRV the pUL37 protein is not essential to replication in cell culture (Klupp et al. 2001). Moreover, a later study has shown that the residues critical for the interaction of VP1-2 and pUL37 can be found more than 150 residues downstream of R5 in HSV-1, in a region which shows significant homology between all α-HV VP1-2s (Mijatov et al. 2007). It has to be noted that at least one more downstream deletion beyond the defined pUL37 binding region in the PRV homologue occurred in that study by Fuchs et al. (2004) which could have compensated for the loss of R5 in some unexplained way and explain why the virus still replicated.

4.6.3 Insertion of a foreign NLS affects viral fitness

The data presented above show that through rescue of the defective ΔNLS virus I was able to isolate and purify recombinant viruses for the chimeras VP1-2.VZV, CMV, EBV and MyoNLS, and another HSV-1 NLS mutant, K428R. All showed varying levels of debilitation on multiple cell lines measured with different techniques which likely indicate a defect during the establishment of infection (likely entry). This effect did not correlate with NLS strength if the whole R1 was exchanged but reduced NLS activity associated with the K428R mutation certainly affected HSV-1 replication. Although there were defects observed in most cell lines tested, the following paragraphs will discuss this defect using Vero cells as a model.

- 129 -

Analysis of NLS function during HSV-1 replication

4.6.3.1 Specific defect in Vero cells

Vero cells turned out to be the worst cell line for propagating these recombinant viruses. This was unexpected since these cells are conventionally used to propagate a variety of viruses since it lacks the type I interferon system. If one considers the defect of the chimeras which is likely due to delayed entry kinetics, this is particularly puzzling. The initial stages of viral infection can be seen as a race to establish the viral programme before the cell can mount an anti-viral response. Due to the lack of the interferon type I system it would be reasonable to assume that slowed kinetics were less of an issue. However, even in the absence of neutralising antibody the VZV chimera barely spread through the monolayer for the multi-step growth analysis indicating a major problem associated with the inserted sequence. The most significant reduction in viral fitness was observed in Vero cells with the VZV chimeric recombinant virus. Detectable levels of ICP4 were delayed by as much as 5 h which is very significant for a virus with a relatively short replication cycle. Such a delay in immediate early gene expression could be caused by slower entry. Since the ΔNLS capsids congregate around the MTOC during early infection (Abaitua et al. 2012) the VZV motif likely fulfils a function allowing docking at the NPC. This would indicate that the process of picking up a nuclear transport factor near or at the NPC is less efficient, in this case not solely because the NLS is weaker but also because of other changes to this region. Alternatively but not mutually exclusively, the required nuclear transport factor is limited in Vero cells and thus prevents mutant capsids from docking at the pore in the same time as wild type virus. This could explain why the K428R recombinant virus was also strongly debilitated in Vero cells (Figure 20B). To test whether presence (or absence) of the required importins could be the limiting factor, expression of these could be tested by Western Blot. Since specific antibodies to each importin have not been completely developed it might be easier to check for expression by quantitative real-time PCR. If this hypothesis were to be true it would also be possible to exogenously express each importin and analyse their capacity to rescue the entry defect, or vice versa knock-down individual importins with siRNA and check if the wild type becomes impaired in a similar or the chimeras lose infectivity completely (if expression of the limiting importin is even further reduced). Besides docking at the nucleus cleavage was suggested to occur while capsids are docked at the NPC. When cells were infected with tsB7 at the non-permissive temperature these capsids can dock at the nucleus but fail to eject their DNA through the NPC when the temperature is switched to the permissive temperature in the presence of a protease inhibitor (Jovasevic et al. 2008). The defect was later traced to a single residue in VP1-2 (Abaitua et al. 2011). A VP1-2 cleavage site is anticipated in a region near R1 (Kattenhorn et al. 2005) but was never confirmed. When you compare the VZV to the

- 130 -

Analysis of NLS function during HSV-1 replication

K428R recombinant, both show smaller plaques, but that of K428R were still >5-fold larger. It could be that the insertion of a foreign R1 could, in addition to affecting binding to nuclear transport factors, impair such a cleavage event making DNA release a very slow process which might be exacerbated in Vero cells more than in other cells. The K428R mutant would in that respect not be affected since it is otherwise wild type sequence. If cleavage were to be an issue this could be tested using a nuclear entry assay. When cells are infected with ultra-high MOI it is possible to see capsids docked at the NPC with EM. This would yield information on how many capsids are docked at the nucleus at any given time and should also show whether the capsids are empty or DNA-filled. One would assume if the above assumptions are true, that fewer capsids reach the nucleus compared to wild type (at any given time) up to a point of saturation and that more mutant capsids would contain DNA. Besides the entry phenotype, virion production also appeared much reduced in Vero cells for all chimeras. This suggests that there might be another defect later on during infection although it could be possible that the viral life was just so much slower that the peak of virion assembly was not reached by 24 hpi. Thus to further assess the defect of the chimeras it would be worth looking at the kinetics appearance of replication compartments and capsids. The Western blot results indicate reduced VP5 content and generally an extended viral life cycle. Thus it appears likely that capsid assembly and maybe also DNA replication will be delayed or much reduced which can be investigated by EM of infected cells (to check for nuclear capsids) or checking how many genomes enter nuclei and progress to become pre-replication compartments. Our group has also optimised a way to analyse the kinetics of viral genome uncoating and DNA replication using ‘Click’ chemistry (unpublished data) which will be a useful tool to delineate the defect of the chimeric viruses in more detail. If entry kinetics were slower, then fewer genomes should enter the nucleus at any given time point compared to wild-type virus. This would also delay the appearance of a large (peak) replication compartment. Where the role lies for the VP1-2 NLS remains to be discovered but it is difficult to reconcile my data and that presented in the literature with a nuclear assembly function. Although it localises to the nucleus and was found on nuclear capsids (D. S. McNabb & Courtney 1992; Abaitua & O’Hare 2008; Henaff et al. 2013) several studies show that it might not be required for nuclear egress (Desai 2000; Möhl et al. 2009). Assembly was not affected using the ΔNLS mutant although nuclear localisation of VP1-2 was abolished (Abaitua et al. 2012). Thus it would be a good idea to check localisation of chimeric VP1-2 though, since altering the NLS could alter its regulation and thus theoretically affect viral assembly. Phenotypes could here be evaluated against the literature by EM which has been done for a variety of VP1-2 mutants already. Of particular interest would be the total number of capsids produced and the appearance of cytoplasmic and perinuclear capsids. Combined with the

- 131 -

Analysis of NLS function during HSV-1 replication aforementioned studies on entry/replication kinetics and ultrastructural investigations of nuclear docking and DNA release this should yield an explanation of why the chimeric viruses might be impaired.

4.6.4 Future directions

The main aim of our group is to solve how VP1-2 facilitates nuclear uncoating and which role the NLS has in that context. In our group the goal was to make HSV-1 research more physiological. Hence cell types that represent more natural host cells were introduced to study the defects of the chimeric viruses. The next steps would be to fully characterise the defect of these viruses and potentially expand the analysis of functional determinants of the HSV-1 R1 region. Of particular interest would be the study of neuronal infection. The HV R1 regions show sub-family specific features and it is intriguing to speculate that some of these arose as an adaptation to the cells they infect. Since most cells tested (except Vero) were still susceptible to the chimeric viruses it would be interesting to expand this study to include neurons. Can the chimeras successfully migrate along the long axons since importin might be required for this process, and if so, can they unload their DNA cargo into the nucleus? Would these chimeras be able to establish latency in an animal model?

- 132 -

Biochemical analysis of VP1-2 binding partners

5 Biochemical analysis of VP1-2 binding partners 5.1 Using the VP1-2 N-terminus to identify relevant nuclear transport factors

Over the last two decades knowledge on how HV capsids unload their cargo into the nucleus has gradually lead to the conclusion that VP1-2 is the major driver of early events including transport and docking at the nuclear pore (see section 1.2 including references). However, the mechanism how capsids dock at the nucleus is only incompletely understood. The accumulation of VP1-2ΔNLS capsids around the MTOC (Abaitua et al. 2012) and rescue assay results imply that one piece of that puzzle could be the recruitment of a nuclear transport factor and that this could well be a conserved function of R1 in all HVs. It would be vital to identify VP1-2 recruited nuclear transport factors to establish a more complete model of nuclear docking of HV capsids. In this chapter I set out to pursue two parallel systems to identify importins recruited by VP1-2 in its soluble and capsid-bound form. The first approach makes use of a N-terminal fragment of VP1-2

(VP1-21-795, NT3, Figure 21B). When exogenously expressed in cells this protein would serve as a bait with which VP1-2 interacting proteins can be co-precipitated and subsequently identified by sensitive MS (see Figure 21A for a flowchart of the proposed work). In combination with a NT3 mutant that lacks the NLS (NT3ΔNLS) nuclear transport factors might be identified by comparing the lists of binding proteins from the separate pulldowns (of NT3 and NT3ΔNLS) since, theoretically, only NT3 should be able to co-precipitate nuclear transport factors. This system may also be useful to answer unrelated questions which could still be affected by the presence of the NLS. During HSV-1 infection VP1-2 was suggested to be cleaved into a product that ends somewhere in the vicinity of R1 (Kattenhorn et al. 2005), that is, either before or after the NLS. The product could thus act either as a nuclear or cytoplasmic protein. NT3 and NT3ΔNLS could mimic these products and the profile of binding partners could yield information on what this cleavage product could do during HSV-1 infection. Herein, I present data on the developed pulldown system and the follow-up of one identified interacting protein.

- 133 -

Biochemical analysis of VP1-2 binding partners

Figure 21 Overview of the strategy to identify binding partners of the N-terminus of VP1-2 and the NLS domain. (A) Flowchart highlighting the steps which were performed to obtain potential binding partners for VP1-2. (B) Constructs used as a basis for the interaction studies. NT3 represents a 795 residue N-terminal fragment of HSV-1 VP1-2 and its nuclear localisation is entirely dependent upon the presence of the NLS. As a control for the proposed MS analysis, a second construct lacking the NLS was also constructed. (C) Retroviruses that enable doxycycline inducible expression of NT3 or NT3ΔNLS were created and used to transduce HEK293 TetON cells which express the rtTA reverse transactivator. The pTight doxycycline regulatable promoter cassette was cloned from pTight (Clontech) with a multi-cloning site (MCS) and inserted into pQ-Flag (Dr. G. Maertens). NT3 or NT3ΔNLS were inserted into the MCS resulting in a retrovirus system that contains a bi-cistronic expression cassette for doxycycline inducible expression of NT3 and puromycin phosphotransferase (selectable marker). Alternatively, the IRES was exchanged for a SV40 promoter to result in the same construct coding for two mRNAs (NT3 and puromycin resistance). pCMV – CMV immediate early promoter, LTR – long terminal repeat, IRES – internal ribosomal entry site, PuroR – puromycin phosphotransferase.

To complement the study of soluble VP1-2 and to validate any identified protein it would be necessary to test whether these would also bind VP1-2 in its capsid-bound form. For this I present data on the optimisation of an in vitro system to identify capsid bound proteins in the context of wild-type

- 134 -

Biochemical analysis of VP1-2 binding partners

VP1-2. The aim of this part was to compare the cellular binding proteins of capsids covered in wild- type and VP1-2ΔNLS by proteomic analysis and Silac. The assay uses capsids that have been treated with detergent to remove the envelope and several conditions (different [salt], reducing environment, cellular proteins) to remove as much of the capsid-associated tegument as to allow binding of relevant cellular proteins. Ultimately, comparing the profile of identified binding proteins precipitated by wild- type and VP1-2ΔNLS mutant capsids should enable the identification of nuclear transport factors that would be recruited during the early entry steps.

5.2 Results 5.2.1 Protein interaction study using the VP1-2 N-terminus 5.2.1.1 Creating stable cell lines expressing VP1-2.NT3

Although NT3 expressed very well in transient expression assays and, more importantly, to higher levels than NT6 I wanted to minimise assay variation due to transfection efficiency and create a cell line with consistently high expression of my bait protein. Since it is possible that the deubiquitinating activity of the USP domain could interfere with cellular function in the long term (i.e. over multiple passages) I opted for a doxycycline inducible system. In this way only when doxycycline was added to the medium would the protein be expressed to high levels. Effectively, in this system exactly the same cell line can be used as positive and negative control (+/- doxycycline). To establish two cell lines expressing NT3 or NT3ΔNLS a retrovirus system was used to transduce a doxycycline regulatable NT3 or NT3ΔNLS open reading frame into HEK293 TetON cells, which express rtTA (protein that activates transcription in response to doxycycline). For this a retrovirus system (kindly provided by Dr G. Maertens) was modified by exchanging the CMV promoter with a pTight promoter comprising seven repeats of the doxycycline transactivator binding site coupled to a minimal CMV promoter (Figure 21C). This, in combination with the reverse transactivator (rtTA), renders the downstream ORF inducible by doxycycline. As part of the cloning process a multi-cloning site was introduced downstream of the pTight promoter to enable insertion of NT3 and potentially other open reading frames. To allow selection this vector contains a puromycin resistance cassette, which was expressed from the same transcript as NT3 through the activity of an IRES. Through puromycin selection it was possible to recover cellular clones, expressing bait proteins to high levels upon induction with doxycycline (Figure 22). To confirm inducible expression of NT3 and NT3ΔNLS, cellular clones for each protein were induced for 19 h or 36 h with different concentrations of doxycycline ranging from 0.001 μg/ml to 1 μg/ml (Figure 22A/B). Without the antibiotic the selected clones did not express detectable

- 135 -

Biochemical analysis of VP1-2 binding partners

NT3/NT3ΔNLS which shows that expression is tightly regulated (Figure 22A/B, lanes 1 and 6 in both blots). For both NT3 and NT3NLS expression was detected at concentrations as low as 0.01 μg/ml and increased to near maximal levels at >0.1 μg/ml at 19 h post-induction (Figure 22A/B, see lanes 4 and 9). There was a considerable amount of shorter protein products that were reproducibly detected with anti-V5 antibody, which detects the N-terminal tag of NT3 (Figure 21B). The observation might indicate cleavage and might be explained by overload of the cell with the protein resulting in detectable N- terminal cleavage products. At high concentration of doxycycline (> 0.1 μg/ml) and more prominently after longer induction, a ladder appears at molecular weights larger than NT3. This could indicate potential ubiquitination which was not removed by the cis-acting USP domain and was only detectable when the protein was vastly overexpressed. This pattern did not change significantly from NT3 to NT3ΔNLS although cellular localisation was altered to a large extent (Figure 22C). To test whether NT3 could practically use the nuclear transport machinery in HEK293 TetON cells, clones for both constructs were induced and analysed by indirect immunofluorescence. It was striking that virtually all cells expressed the bait protein while control cells not treated with the antibiotic did not show any signal in the FITC channel if imaged with the same microscopy settings for gain and exposure (Figure 22C). NT3 showed moderate to strong nuclear accumulation in the majority of cells (Figure 22C) which was virtually abolished with the deletion of the NLS. It has to be noted that NT3 still accumulated in the cytoplasm of cells which likely could be due to overloading the cell with nuclear transport cargo. Nonetheless these data imply that the selected clones express proteins that could be used to potentially identify nuclear transport factors.

- 136 -

Biochemical analysis of VP1-2 binding partners

Figure 22 Characterisation of transduced HEK293 TetON cell clones. (A and B) Selected clones from NT3 wild-type (A) and ΔNLS (B) were grown in collagen-coated 6-well plates and induced the next day for either 19h (lanes 1-5) or 36h (lanes 6-10) using increasing concentrations of doxycycline (0, 0.001 μg/ml, 0.01 μg/ml, 0.1 μg/ml or 1 μg/ml). Proteins were extracted by scraping cells into Laemmli buffer and NT3 was detected by Western blot using anti-V5 antibody. γ-tubulin was used as a loading control. (C) Immunofluorescence studies of one selected cell clone for each construct is shown. NT3 or NT3ΔNLS were induced with 1 μg/ml doxycycline for 19 h and their presence was detected using anti-V5 antibody. Nuclei were counterstained with DRAQ5. NT3ΔNLS exhibited exclusively cytoplasmic localisation whereas NT3 was also detectable in the nucleus and cytoplasm.

- 137 -

Biochemical analysis of VP1-2 binding partners

5.2.1.2 Identification of VP1-2 interacting proteins: Part I

To find interaction partners two cellular clones inducibly expressing NT3 were initially compared to the NT3ΔNLS clone using immunoprecipitation. Western blot analysis showed that all NT3 and NT3ΔNLS clones expressed a protein of the same size although NT3 clone 15 and NT3NLS appeared to express slightly better than NT3 clone 4 at the conditions tested (Figure 23A, compare lane 2-4). The control parental HEK293 TetON cell line showed no detectable band at the expected size (Figure 23A, lane 1). Moreover, NT3 species (lanes 11-14) were precipitated only in NT3 and NT3ΔNLS clones but not in the parental control. Conversely, the cytoplasmic protein γ-tubulin was not enriched in any IP sample. It has to be noted that significant amounts of NT3 were still present in the unbound, soluble fraction after the precipitation step (Figure 23A, lanes 8-10) indicating poor recovery which needs to be improved to gain sensitivity. Protein profiles of the same samples analysed on a silver stained SDS- PA gel (Figure 23B) showed strong bands at the expected size of NT3 which were not present in the parental control. There were some significant contaminating, non-specific bands in the parental control IP (examples marked with green arrows). These could not be removed by pre-clearing of the cellular lysates with protein-A bearing beads which were used to immobilise the anti-V5 antibody. Since these bands re-appeared in repeat experiments it is likely that they get co-precipitated non- specifically by anti-V5 antibody. The banding pattern of the NT3 sample was reproducible with two NT3 clones and over multiple experiments for the same NT3ΔNLS clone. The strong bands above (60 kDa) and below (40 kDa) the immunoglobulin (Ig) heavy chain likely match the N-terminal breakdown products also seen in the Western blot (Figure 22) and would thus be of no significance. Other bands that were present both in NT3 and NT3NLS but not in the control sample could be interacting proteins of any region of NT3; the N-terminal USP, the region after the USP (including the NLS) or the C-terminal part. To discover transport factors, lanes 2 or 3 (NT3) have to be compared to lane 4 (NT3ΔNLS) (Figure 23B). Several bands were reduced in staining or failed to show up completely in the NT3ΔNLS sample compared to wild-type NT3 sample (Figure 23B, lane 4, red arrow heads). Two bands between 70-80 kDa (black arrowheads), which were absent from NT3ΔNLS, were excised, tryptically digested and analysed by tandem MS. In those bands six proteins of approximately the correct size were at least four-fold enriched in the NT3 compared to the NT3ΔNLS sample (see Figure 23C). The proteins identified are involved in nuclear functions including transcription (DDX3X, DDX17, PABP1) and repair of damaged DNA (DTX3L) and have not previously been identified as VP1-2 interacting proteins. Their enrichment in the NT3 sample might not directly reflect the deletion of the NLS but rather the change in localisation of NT3 compared to NT3ΔNLS. This band excision analysis represented an initial validation that proteins enriched in the NT3 compared to the NT3ΔNLS sample could be identified by MS. This result

- 138 -

Biochemical analysis of VP1-2 binding partners indicated that it might be a feasible strategy and thus I proceeded to the next stage which involved an unbiased MS screen of all proteins precipitated by NT3 or NT3ΔNLS.

Figure 23 Small scale immunoprecipitation of NT3 and NT3ΔNLS to identify VP1-2 interacting proteins. (A) Western blot showing the presence of NT3 and NT3ΔNLS in cellular lysates and the IP fraction (detected with anti-V5 antibody). (B) Silver stained SDS-PA gel of the same IP samples after elution from protein A beads using Laemmli buffer. Green arrowheads show examples of non-specifically precipitated proteins. Red arrowheads show areas where bands were less intense or absent in the NT3ΔNLS sample compared to NT3. Black arrowheads show bands that were excised for tryptic digestion and subsequent mass spectrometric analysis. (C) Selected proteins identified in the excised bands by MS. Only proteins that were enriched at least four-fold in NT3 versus NT3ΔNLS and that were of approximately the correct size were included in the list. Values for NT3 (column 2) and NT3ΔNLS (column 3) show absolute abundance with a difference of 1 unit between NT3 and NT3ΔNLS indicating two-fold enrichment.

- 139 -

Biochemical analysis of VP1-2 binding partners

5.2.1.3 Identification of VP1-2 interacting proteins: Part II

To get an overview of which proteins the various bands in the NT3 or NT3ΔNLS samples represented (see Figure 23B) the IP was repeated on a larger scale and all co-precipitated proteins were directly subjected to tryptic digestion and the resulting peptide mixtures compared by mass spectrometric analysis. Western blot analysis showed enrichment of NT3 in the IP fraction which was subjected to tryptic digestion (Figure 24A, lanes 8 and 9, similar background bands as before, full blot not shown). Again no γ-tubulin was found in the IP fraction and a reasonably high amount of NT3 remained in the soluble fraction after the precipitation step (lanes 6 and 7). The MS results revealed highly similar levels of precipitated NT3 and NT3ΔNLS (NT3/NT3ΔNLS ratio 0.999) in either sample, with a total of 43 and 35 peptides identified for NT3 and NT3NLS, respectively. This was corroborated by comparable band intensity in the IP fraction (see Figure 24A, lanes 8-9). Overall, NT3 (and NT3ΔNLS) showed more than 8000-fold enrichment in the expressing cell clones compared to the parental control, which was deemed acceptable to proceed with the analysis of MS results.

Figure 24 NT3 and NT3ΔNLS IPs and summary of co-precipitated proteins identified by MS. (A) Western blot showing presence of NT3 and NT3ΔNLS in samples destined for MS. There was some degradation detected in Total and Input samples but mostly full length NT3 was precipitated with anti-V5 antibody. γ-tubulin was used as loading control. (B) Schematic overview of proteins that were >4 fold enriched in the NT3/NT3ΔNLS samples compared to the parental control. Of the 199 proteins that were found, 162 and 93 were enriched in NT3 (red, category I) and NT3ΔNLS (green, category II), respectively. 56 proteins were found in both samples (brown). The 199 proteins were further sub-

- 140 -

Biochemical analysis of VP1-2 binding partners divided. Category III includes 99 proteins that were >4 fold enriched in NT3 compared to NT3ΔNLS. Category IV comprises 32 proteins that were enriched in NT3ΔNLS compared to NT3. Category V includes 54 proteins that were found enriched in both NT3 and NT3ΔNLS compared to control (and less than 4 fold different between NT3 and NT3ΔNLS). 2 proteins from the overlap were assigned to category III due to >4 fold enrichment in NT3>NT3ΔNLS. 9 proteins from Cat. I and 5 from Cat. II were excluded as the co-precipitated protein was only enriched in one sample compared to control but the difference between the bait samples was less than four fold.

The MS results were filtered to remove any proteins that could typically contaminate MS (e.g. trypsin from the digestion, albumin, etc.) and to remove proteins that were poorly enriched in both bait protein samples (i.e. less than four-fold enrichment compared to parental control). A list of 199 proteins remained (Figure 24B). Of these, 162 and 93 proteins were found to bind NT3 or NT3NLS, respectively, with 56 proteins represented in both samples. Of the 162 proteins that bound NT3 better than the beads (parental control) 99 were additionally at least four-fold more abundant compared to the NT3ΔNLS sample (Figure 24B, NT3 > NT3ΔNLS, category III). Of the 93 proteins identified in the NT3ΔNLS sample (compared to control) 32 proteins were additionally at least four fold enriched compared to the NT3 sample (Figure 24B, NT3ΔNLS > NT3, category IV). 54 of the remaining proteins were at least four fold enriched in both NT3 and NT3NLS samples compared to the parental control (Figure 24B, category V), but showed a less than four-fold difference between NT3 and NT3ΔNLS samples.

5.2.1.4 Localisation and functions of proteins that were enriched compared to the parental control (categories I and II)

It could be expected that proteins that bind a nuclear bait would naturally also localise to this compartment and vice versa for proteins binding a cytoplasmic bait. This data could then be taken as an internal validation of the pulldown and aid the categorisation of identified proteins. For this experimentally confirmed localisation data was gathered manually from UniprotKB (and GeneCards if UniprotKB did not contain any annotation). While some of the 199 proteins pass through multiple compartments, at least 67% of the 162 proteins enriched in NT3 over control (Figure 25A, category I) localise to the nucleus at some stage compared to only 56% of the 93 proteins that bound NT3ΔNLS better than control (Figure 25B, category II). This manual analysis compared well with the data obtained with the automated annotation tool DAVID (Huang et al. 2007), where at least 60% of the 162 proteins enriched in the NT3 sample versus control (category I, Fisher score <5x10-26) were categorised as nuclear proteins compared to 52% of the 93 proteins enriched in the NT3ΔNLS sample versus control (Fisher score <1.5x10-8). The low Fisher scores indicate statistically significant enrichment of nuclear proteins (more than expected by chance in a sample set) in both sample sets, although the enrichment in the NT3 sample (category I) was expectedly much higher. Interestingly, 32

- 141 -

Biochemical analysis of VP1-2 binding partners

NT3 binding proteins (category I) were also found in nucleoli in some experimental settings, which suggests that NT3 might localise to this compartment. In agreement with this theory R5 is predicted to contain a nucleolar localisation signal and could explain sub-nuclear compartmentalisation of GFP-R2 (see Figure 9B, panel III).

Figure 25 Simplified overview of annotated localisation and functions of NT3 binding proteins Localisation of proteins that were enriched at least four-fold in the NT3 sample (A and C, category I) or NT3ΔNLS sample (B and D, category II) versus the parental control are shown. Annotations were collated mainly from UniProtDB but also GeneCards if no annotation was found. Some proteins localise to multiple compartments or have multiple annotated functions. Cytoplasmic includes other compartments (ER, Golgi, mitochondriae, etc.). The top five functions are also shown as percentage of all proteins from that category.

It could be expected that importins would be found in both category I and III (Figure 24B) since NT3 contains the NLS. However, no nuclear transport factors were identified in the NT3 sample. This implies that this type of assay might not be the right one or that it requires modification. Nonetheless this assay could yield valuable information on functions of another part of NT3; the USP, which gets

- 142 -

Biochemical analysis of VP1-2 binding partners liberated by a proteolytic event during infection (Kattenhorn et al. 2005). The putative cleavage site in VP1-2 is still unknown but lies somewhere around R1 both NT3 and NT3ΔNLS. Additionally, VP1-2 can be found both in the nucleus and in the cytoplasm in its full length form (D S McNabb & Courtney 1992; Abaitua & O’Hare 2008). This could effectively mean that a free USP domain can affect targets either in the nucleus (if the cleavage site is after R1) or the cytoplasm (if the cleavage event occurs before the NLS). The USP can act on VP1-2 to modify its own stability (Bolstad et al. 2011) and another report exist of the VP1-2 de-ubiquitinating activity modulating the immune response to HSV-1 infection (Wang et al. 2013). Thus the NT3 and NT3NLS system could potentially lead to the identification of VP1-2 interacting proteins and a more complete understanding of how VP1-2 fulfils some of its functions. In fact in another study (Calistri et al. 2014) an interaction of VP1-2 and a component of the ESCRT pathway was found using a NT3-like fragment. With these precedents I believe that following up some of the identified proteins could give rise to new ideas in how HSV-1, via VP1-2 or the USP, modulates cellular behaviour. Of particular interest is the group of proteins that were enriched in both samples compared to the control (Figure 24B, category V) since they bound both bait proteins (similar to a biological replicate). To gain an insight into what the sample proteins could be doing during an infection, annotated molecular functions were manually gathered from UniprotKB (and GeneCards if no annotation was found in Uniprot) for proteins that bound NT3 and/or NT3ΔNLS (Figure 25C/D). These were divided into 13 functional groups. A great majority of the 199 proteins have roles handling RNA with approximately 27% of all proteins presumably involved in transcription or mRNA metabolism. An additional 11% function in other RNA metabolism, such as RNA binding which could be involved in immune recognition, tRNA and miRNA metabolism. Besides this, 10% of proteins were shown to be involved in translation and ribosome function and, in fact, many of the identified proteins play prominent roles at every level in the chain of protein expression which could be modulated to keep up protein production and gear this process towards expression of HSV-1 proteins. The second largest functional grouping relates to cell cycle and DNA replication. 13% of proteins that were enriched compared to control (of all 199 proteins) were shown to be implicated in these functions which may contribute to modulation of the replicative machinery to allow replication of viral genetic material. A small proportion (6%) of proteins are associated with transport functions which could be interesting since during an infection HSV-1 capsids have to navigate the cytoplasm to reach the nucleus or the place of virion assembly. Unexpectedly, the only identified nuclear transport factor (importin-8) was found in the NT3NLS data sets. This implies that, despite deletion of the whole basic patch in R4, importins may still bind to NT3 without fulfilling their transport function.

- 143 -

Biochemical analysis of VP1-2 binding partners

A closer look at the raw data, however, reveals that additional nuclear import receptors were identified. On the one hand, importin-9 which also bound to NT3ΔNLS. On the other hand Importins β1, and importins 4, 5 and 7, and importins alpha 1, 3, 4, 5 and 7 were also precipitated in the parental HEK293 TetON control. In fact, most nuclear transport factors, except for importin-9, were found to bind best to the beads in the parental control sample, which did not contain any bait protein. This implies that in spite of pre-clearing over protein A beads these proteins still bound to beads coated with anti-V5 antibody which presents a substantial problem if the aim was to identify these factors. To get at functions of potential targets of the USP domain in the context of NT3 (or NLS) the functions of all identified proteins in categories I and II were analysed. In categories I and II, which represent NT3 and NT3NLS binding proteins (Figure 25C/D), respectively, the former showed comparative enrichment for proteins involved in cell cycle (15% in the NT3 vs 9% in the NT3ΔNLS sample), transcription (16% vs 8%) and DNA repair and genome structure (8% vs 5%), all proteins that likely function in the nucleus. Conversely, proteins that preferentially bound NT3NLS are involved in metabolism (9% in the NT3ΔNLS vs 3% in the NT3 sample), transport (12% vs 8%) and protein folding (chaperones; 11% vs 4%), which likely function in the extra-nuclear compartments A select set of proteins divided into categories III to V are listed in Table 8 (proteins enriched in the NT3 compared to the NT3ΔNLS sample), Table 9 (proteins enriched in the NT3ΔNLS compared to the NT3 sample) and Table 10 (enriched in both samples compared to the control sample). A particularly interesting duo of binding proteins were DTX3L and its interaction partner PARP9. Interestingly, these two proteins are thought to form a complex which is implicated in the DNA damage response (Yan et al. 2009; Yan et al. 2013). Both were enriched in the NT3 and NT3ΔNLS samples compared to control (Table 8 and Table 10). DTX3L was also identified in the gel slices excised from the initial silver stained SDSPA-gel (Figure 23C). In absolute terms the levels of DTX3L and PARP9 precipitated by NT3 were almost identical as judged from the peptide intensity score (compare Table 8 and Table 10) but NT3ΔNLS appeared to bind PARP9 slightly better than DTX3L. These data suggest that NT3 potentially co-precipitated these proteins as a complex. The increased level of PARP9 compared to DTX3L in the NT3ΔNLS indicates that DTX3L might not be precipitated through a direct interaction with NT3 but rather via PARP9, although this cannot be said with full confidence (more experiments required).

- 144 -

Biochemical analysis of VP1-2 binding partners

Table 8 Selected Proteins that bound NT3 >four-fold better than NT3ΔNLS and parental control (Category III). Absolute intensity given as log2.with a change of 1 unit equalling a doubling in abundance. Experimental details are given in Figure 24. The full list of enriched proteins can be found in 12: Appendix 4.

intensity unique peptides PEP MS/MS # NT3 MW Name Gene Function(s) important for HSV-1? References NT3 parental score Count NT3 NT3NLS control ΔNLS (Miyazaki upregulated 4.8x in retinal endothelial cells; et al. 2011; E3 ubiquitin-protein ligase 1 30,075 27,4888 0 6E-174 34 18 15 0 84 DTX3L Ubiquitination recruitment 53BP1 to DNA damage sites; stimulates Yan et al. DTX3L NHEJ which is detrimental to HSV-1 infection 2009; Yan et al. 2013) 6,29E- Mediator of RNA polymerase required for HSV, HCMV and slightly less for VZV (Griffiths et 2 28,348 24,2055 23,0033 14 15 4 3 161 MED14 transcription 86 II transcription subunit 14 infection (siRNA) al. 2013) miRNA 3 23,143 0 0 3,2E-09 4 2 2 2 219 Endoribonuclease Dicer DICER1 No reference found metabolism (Griffiths et al. 2013; modulation of LIM-HD function; involved in neuronal 2,92E- E3 ubiquitin-protein ligase Patel et al. 4 24,645 0 0 3 3 0 0 69 RLIM Ubiquitination development; LIM knockdown blocks HSV-1 09 RLIM 1998; infection; LIM-HD factor synergises with NFkB Rascle et al. 2009) (Meggio & protein kinase, at least six HSV-1 proteins phosphorylated by CK2; 6 26,483 24,3834 0 4,6E-14 3 3 2 1 25 Casein kinase II subunit beta CSNK2B Pinna multiple pathways EBNA-2 and ZEBRA of EBV 2003) 4,61E- DNA polymerase delta ubiquitinated due to DNA damage; DDR upregulated (Lee et al. 7 25,459 0 0 3 3 1 1 50 POLD2 cell cycle 12 subunit 2 in HSV-1 infected cells 2014) 5,54E- General transcription factor (Griffiths et 8 25,73 0 0 3 3 1 1 101 GTF3C2 transcription required for HSV infection (siRNA) 13 3C polypeptide 2 al. 2013) 2,06E- E3 ubiquitin-protein ligase potentially associated with HSV-1 susceptibility in (Thompson 9 31,293 0 0 2 2 1 1 201 LTN1 Ubiquitination 06 listerin mice et al. 2014) inhibits HSV-1 infection (siRNA); facilitates TNF 9,19E- (Griffiths et 10 22,807 0 0 2 2 1 0 40 Modulator of apoptosis 1 MOAP1 apoptosis mediated apoptosis; K63 ubiquitination at K278 04 al. 2013) appears to promote cell death (Schang et 3,92E- Cyclin-dependent kinase 2 CDK2 protein kinase, cell CDK2 activated during HSV-1 infection, Cdk inhbition al. 1999; 11 25,013 0 0 2 2 0 0 27 13 Cyclin-dependent kinase 3 CDK3 cycle blocks HSV-1 infection Schang et al. 1998) 5,10E- Cyclin-dependent kinase (Griffiths et 12 26,633 0 0 2 2 1 1 14 CDKN2A cell cycle appears to inhibit HSV-1 replication (siRNA) 11 inhibitor 2A, isoform 4 al. 2013)

- 145 -

Biochemical analysis of VP1-2 binding partners

intensity unique peptides PEP MS/MS # NT3 MW Name Gene Function(s) important for HSV-1? References NT3 parental score Count NT3 NT3NLS control ΔNLS 3,10E- Cell division cycle 7-related protein kinase, cell promotes cell survival in presence of DNA damage, (Suzuki et 13 23,477 0 0 2 2 0 0 61 CDC7 04 protein kinase cycle promotes DNA replication al. 2012) 2,69E- Ras GTPase-activating-like small GTPase, cell 14 24,374 0 0 2 2 1 1 179 IQGAP3 15 protein IQGAP3 cycle, signalling in vesicles with activated STING to effect innate 8,64E- Exocyst complex component (Ishikawa 15 26,169 0 0 2 2 0 1 104 EXOC2 exocytosis response to DNA (IFN-β); STING is anti-HSV-1 04 2 et al. 2009) effector mRNA required for HSV infection (siRNA); also important for (Griffiths et 16 23,827 0 0 8,4E-06 2 2 0 0 75 Symplekin SYMPK metabolism West Nile Virus infection al. 2013) (Burgdorf 2,86E- apoptosis, protein binds Tsg101; ubiquitinated/degraded during et al. 2004; 17 24,032 0 0 1 2 1 1 63 Protein AATF AATF 10 kinase apoptosis De Nicola et al. 2014) Cyclin-dependent kinase CDKN2A (Griffiths et 18 23,544 0 0 1,7E-06 1 2 1 0 11 inhibitor 2A isoforms 1/2/3 cell cycle inhibits HSV-1 infection (siRNA) CDKN2B al. 2013) Cdk4 inhibitor B (McElwee 3,46E- plus-end directed capsid transport, dystonin = NLS et al. 2013; 19 23,439 0 0 1 2 1 1 616 Dystonin DST cytoskeletal 05 phenotype Pasdeloup et al. 2012)

- 146 -

Biochemical analysis of VP1-2 binding partners

Table 9 Selected proteins that bound NT3ΔNLS >4-fold more than both NT3 and control (Category IV). Absolute intensity given as log2.with a change of 1 unit equalling a doubling in abundance. Experimental details are given in Figure 24. The full list of enriched proteins can be found in Appendix 5: Category IV (proteins at least four fold enriched)

intensity PEP MS/MS unique peptides # MW Name Gene Function(s) important for HSV-1? References NT3 NT3 ΔNLS parental score Count NT3 NT3NLS Parental binds UL20 (Y2H) which is required for membrane 9,31E- (Griffiths et 1 0 26,6598 0 7 1 4 0 11 Protein S100-A7 S100A7 immunity fusion during assembly, upregu-lated during 14 al. 2013) inflammation/wound healing Prostaglandin E synthase 2 (Kurane et 2,69E- metabolism, COX-2 upregulated in neurons upon reactivation, 2 0 24,275 0 3 0 3 1 42 Prostaglandin E synthase 2 PTGES2 al. 1984; Liu 06 immunity COX-2 inhibition reduced reactivation in vitro truncated form et al. 2014) 1,12E- nuclear transport 4 0 24,2396 0 2 1 3 2 120 Importin-8 IPO8 05 factor 6,57E- Nuclear pore glycoprotein nuclear pore 5 0 23,1143 0 2 1 2 1 46 NUP62 05 p62 protein tight junctions protect HSV-1 from neutralising (Abaitua et 6 0 23,6455 0 0,00102 2 1 2 1 134 Tight junction protein ZO-2 TJP2 cell adhesion antibodies (cell-cell spread) in HaCaT cells al. 2013)

- 147 -

Biochemical analysis of VP1-2 binding partners

Table 10 Proteins that were >4-fold enriched in both NT3 and NT3ΔNLS samples compared to parental (Category V). This table includes proteins that are less than four fold different between NT3 and NT3ΔNLS. Absolute intensity given as log2.with a change of 1 unit equalling a doubling in abundance. Experimental details are given in Figure 24. The full list of enriched proteins can be found in Appendix 6: Category V (proteins at least four fold enriched)

intensity unique peptides PEP MS/MS # NT3 paren MW Name Gene Function(s) important for HSV-1? References NT3 score Count NT3 NT3NLS parental ΔNLS tal (Krenciute et al. 2013; UBL4A-GERT4-BAG6 complex mediates BRCA1 Large proline-rich protein Thierry- 1 32,719 30,7855 26,4064 0 106 38 28 10 119 BAG6 Chaperone recruitment and apoptosis in response to DNA BAG6 Mieg & damage Thierry- Mieg 2006) (Kim et al. retrograde endosomal transport; Tip-VPS35 Vacuolar protein sorting- 2010; 2 30,833 30,1861 23,2049 4E-209 46 21 17 3 92 VPS35 transport interaction during HVS infection downregulates associated protein 35 Kingston et CD4 al. 2011) E3 ubiquitin-protein ligase Ubiquitination, 3 30,157 30,5403 27,4961 9E-189 44 18 15 12 35 STUB1 CHIP DNA repair (Miyazaki et Poly [ADP-ribose] polymerase DNA repair, 4.2x upregulated during HSV-1 infection of corneal al. 2011; 4 30,198 28,3735 0 3E-169 43 23 17 1 96 PARP9 9 immunity endothelial cells; binds DTX3L Takeyama et al. 2003) (Griffiths et 5 28,451 27,5324 24,5152 2,4E-36 17 8 7 3 18 Ubiquitin-like protein 4A UBL4A transport required for HSV infection (siRNA) al. 2013) Vacuolar protein sorting- (Kim et al. 6 28,02 27,7688 0 3,3E-50 17 10 6 0 39 VPS26B transport in complex with VPS35 (see above) associated protein 26B 2010) (Thierry- Mieg & 7 28,473 27,2566 25,8036 1E-55 15 8 7 5 39 ATPase ASNA1 ASNA1 transport binds UBL4A Thierry- Mieg 2006) (Krenciute et al. 2013; Golgi to ER traffic protein 4 Thierry- 8 27,337 26,1278 0 5,4E-44 15 6 5 1 37 GET4 transport binds UBL4A, see BAG6 homolog Mieg & Thierry- Mieg 2006) (Griffiths et 9 27,096 25,3863 24,2501 6E-40 13 7 4 2 28 Myeloid leukemia factor 2 MLF2 not annotated inhibits HSV infection in Hela cells al. 2013)

- 148 -

Biochemical analysis of VP1-2 binding partners

intensity unique peptides PEP MS/MS # NT3 paren MW Name Gene Function(s) important for HSV-1? References NT3 score Count NT3 NT3NLS parental ΔNLS tal U4/U6.U5 tri-snRNP- mRNA (Griffiths et 10 22,795 23,1311 0 9,1E-05 5 2 2 2 90 SART1 required for HSV infection (siRNA) associated protein 1 aumetabolism al. 2013) 3,28E- BRCA1-associated ATM (Aglipay et 11 25,021 23,4432 22,9097 5 4 2 2 88 BRAT1 DNA repair Kinase of ATM in DNA repair 17 activator 1 al. 2006) Nucleolar and coiled-body (Griffiths et 12 24,405 23,9568 0 1,1E-05 4 3 3 2 74 NOLC1 cell cycle required for HSV infection (siRNA) phosphoprotein 1 al. 2013) (Yamaguchi Ubiquitin carboxyl-terminal modulates NFkB activation by targeting IkB; et al. 2007; 13 24,057 23,1927 0 1,4E-11 3 3 2 1 105 hydrolase Ubiquitin carboxyl- USP11 Ubiquitination facilitates survival in repsonse to DNA damage Schoenfeld terminal hydrolase 11 et al. 2004) mRNA DNA-directed RNA polymerase required for HSV infection (siRNA); ubiquitinated (Griffiths et 14 24,155 25,1036 0 1,3E-10 3 2 2 1 16 POLR2D metabolism, II subunit RPB4 during DNA damage al. 2013) translation Vacuolar protein sorting- VPS4A/B cell cycle, (Crump et 15 23,974 23,9065 0 0,00068 3 2 2 1 49 required for HSV-1 envelopment associated protein 4A /B FIGNL1 transport al. 2007)

- 149 -

Biochemical analysis of VP1-2 binding partners

5.2.1.5 NT3 interacts with DTX3L in HEK293 cells

To follow up on the candidate VP1-2 binding protein DTX3L, which was identified by MS in the excised band from the NT3 sample, and in both samples from the large scale IP (Figure 23C, Table 8) the previously successful IP approach was repeated to precipitate NT3 (see Figure 26A-C). Briefly, HEK293TetON cells expressing either NT3 or NT3ΔNLS were induced with doxycycline before the cells were lysed. From separate sets of lysates either NT3, NT3ΔNLS (using anti-V5 antibody) or DTX3L (using a DTX3L specific antibody) were precipitated and the resulting precipitates analysed for the presence of DTX3L or NT3/NT3ΔNLS (Figure 26). Again, NT3 and NT3ΔNLS were precipitated in good amounts but much of the bait protein from the input fraction was not precipitated which was consistent with previous results (Figure 26, lanes 11/12). Both versions of NT3 were precipitated at reasonable and similar levels (lanes 5/6) and when the same blot was probed with a polyclonal antibody to DTX3L, the protein appeared to be present in both NT3 samples, although it seemed that NT3ΔNLS bound it less well (lanes 2/3). This was in agreement with the MS results indicating at least four fold less DTX3L co-precipitated with NT3ΔNLS compared to NT3 (Table 8). There was no DTX3L precipitated in the parental control (lane 1) despite the levels in the input fraction of this sample being comparable to the NT3 and NT3ΔNLS samples (compare lanes 7-12 for input and unbound). This also corroborated the lack of detection of DTX3L in the MS control sample. Intriguingly, the level of DTX3L appeared slightly increased in the cells that express NT3 (Figure 26, lanes 8, 11, 20 and 23) compared to both the control (lanes 7, 10, 19 and 22) and NT3ΔNLS (lanes 9, 12, 21 and 24) expressing cells. This effect on DTX3L levels appeared reproducibly in two biological replicates (i.e. the anti-V5 and anti-DTX3L IPs were performed with independent cellular lysates). The reverse IP (anti-DTX3L; Figure 26D/E) using a polyclonal antibody to DTX3L appeared to co- precipitate both NT3 and NT3ΔNLS (lanes 17/18), although the anti-DTX3L antiserum only poorly precipitated the antigen and additionally gave rise to multiple bands (Figure 26D). This, in combination with some technical difficulties gave rise to ambiguous results for the reverse IP since a co-migrating band corresponding to NT3 in the negative parental control sample, which theoretically did not contain NT3, was detected (Figure 26E, lane 16, red asterisk).

- 150 -

Biochemical analysis of VP1-2 binding partners

Figure 26 Western blot images showing the results from reciprocal co-immunoprecipitations of NT3 (A-C) and DTX3L (D-F). Either NT3 or DTX3L were precipitated from 2x107 dox-induced HEK293 cell clones (and parental control containing neither NT3 nor NT3ΔNLS). IP eluates (IP against V5 or DTX3L) were blotted and membranes probed for presence of NT3 (in B the V5 epitope and in E an anti-USP antibody were used) or DTX3L were detected on the same blot (Licor two colour imaging system). Input and unbound (soluble left after precipitation) fractions correspond to 1/200th of the total while one quarter of the IP fraction was loaded. Red asterisk demarcates a band detected with anti-USP antibody in the parental control.

- 151 -

Biochemical analysis of VP1-2 binding partners

5.2.1.6 NT3 abolishes clustering of 53BP1 in response to the DNA damaging agent doxorubicin

One of the roles of DTX3L (and PARP9) is to aid recruitment of factors, such as 53BP1 (p53 binding protein), via ubiquitination (Yan et al. 2013), to sites of dsDNA breaks (Huen & Chen 2008). 53BP1 is thought to favour non-homologous end joining (NHEJ), which is a detrimental process during HSV-1 infection (Taylor & Knipe 2004) and may be forming part of an interferon induced immune response against HSV-1 (Juszczynski et al. 2006; Lilley et al. 2011). To see if HSV-1 VP1-2 could modulate DNA repair mechanisms, 53BP1 localisation in the NT3 expressing cells in response to DNA damaging agents like doxorubicin could be analysed (e.g. 53BP-1 foci formation). For this the previously described clone expressing doxycycline inducible NT3, was grown with or without doxycycline for 24 h (to induce expression of NT3) prior to inducing DNA damage through treatment with 100nM doxorubicin for 22 h (Figure 27A). In parental cells 53BP1 foci could be seen in untreated cells (panel C1). This was not affected by treatment with doxycycline (Figure 27A, compare panels A1 no doxycycline, B1 with doxycycline). By 22 h of doxorubicin treatment virtually all non- induced cells showed 53BP1 foci (Figure 27A, panels A1, B1) indicative of an active DDR. Some cells in panels C1 showed few or no foci. Some DNA damage inherently occurs during DNA replication thus few, sometimes quite large foci (red arrowheads) could be seen even in untreated cells. Since doxorubicin induced damage is maximised during actual DNA replication (Gewirtz 1999) this indicates that cells without foci (green arrowheads) probably were not actively synthesising DNA. In non-induced NT3 expressing cells (Figure 27A, panels A2) similar 53BP1 foci were observed as for parental cells (Figure 27A, panels A1, B1). Strikingly, these 53BP1 foci were virtually absent in the presence of NT3 after 22 h of DNA damage (B2). Foci disruption led to diffuse nuclear staining of 53BP1 (Figure 27A, panels B2, yellow arrowheads). Unsurprisingly, NT3 also affected the basal DDR as judged by absence of 53BP1 foci (Figure 27A, panel C3). To confirm the above result, the assay was repeated in U2OS transfected with NT3. DNA damage was induced with 100 nM doxorubicin 24 h post-transfection and after 22 h of incubation 53BP1 foci were quantified. Cells with more than 10 nuclear 53BP1 foci were regarded as positive for DNA damage since cells not treated with doxorubicin generally showed fewer than 10 foci per nucleus while at least 80 % of cells exposed to doxorubicin showed in excess of 10 foci (Figure 27B, bars 1 and 2). NT3 reduced the number of cells positive for DNA damage foci to around 25 % (Figure 27B, bar 4) while an unrelated exogenously expressed nuclear protein, CREB-H, did not affect foci formation (Figure 27B, bar 3). Unsurprisingly, cells expressing more NT3 showed fewer 53BP1 foci reducing the number of positive cells to baseline (Figure 27B, bar 6). This reduction in 53BP1 recruitment was dependent on USP activity, since a USP-inactive mutant (Figure 27B, bar 5 and 7, NT3C65A) did not reduce the number of

- 152 -

Biochemical analysis of VP1-2 binding partners cells showing more than 10 foci per nucleus. Taken together, these data suggest a role for the VP1-2 USP domain in modulating the DDR, probably facilitated via its interaction with DTX3L.

Figure 27 53BP1 foci formation in NT3 expressing cells after doxorubicin treatment. (A) Previously described NT3 expressing HEK cells were induced with doxycycline for 24 h and subsequently treated with DNA damaging reagents (doxorubicin 100nM) for 22 h. The cellular response was measured by staining for 53BP1

- 153 -

Biochemical analysis of VP1-2 binding partners

(MAB3802) which is thought to form foci in the nucleus. The red arrowhead demarcates a representative untreated nucleus with normal, replication-induced 53BP1-containing DNA damage foci; the green arrowhead shows a representative nucleus without 53BP1 foci (i.e. cells that have not synthesised DNA during doxorubicin treatment); the yellow arrowhead shows a representative nucleus with diffuse 53BP1 localisation lacking obvious foci. (B) U2OS cells transiently expressing NT3, its USP-inactive mutant NT3.C65A or the unrelated transcription factor CREB-H were exposed to doxorubicin (100 nM) for 22 h and 53BP1 foci formation was quantified. Cells with > 10 foci per nucleus were regarded as positive of DNA damage since less than 5 % of cells showed > 10 foci per nucleus in the absence of DNA damaging agents. Cells positive for DNA damage generally showed between 10 and 60 foci with a mean of approximately 30. High expression of NT3 was discriminated by applying a threshold using the Fiji software. More details on quantification can be found in the materials section.

- 154 -

Biochemical analysis of VP1-2 binding partners

5.2.2 Development of a system to identify cellular binding partners of incoming viral capsids

To complement the investigation of interaction partners of soluble VP1-2 and the NLS, I attempted to develop a system to delineate the interactome of incoming capsids. For the purposes of this analysis it was important to mimic the features of capsids on the way to the NPC, that is, during the first 5-30 min after infection. During that time virions would have lost their envelope either through fusion at the plasma membrane or endosomes. Subsequently, capsids are sequentially stripped off most of the tegument structures except for, theoretically, VP1-2, pUL37 and pUS3 (see 1.1.5). To obtain capsids fulfilling these requirements three different methods could principally be employed.

5.2.2.1 Possible approaches to identify capsid-binding nuclear transport factors (i) nuclear capsid preparation

The first approach would obtain newly assembled capsids from nuclear preparations as these could easily yield high amounts of material. These could be used as an affinity matrix for potential binding partners. However, these capsids may not look like the ones during entry (see 1.2.3). Conflicting evidence exists as to whether VP1-2 is recruited to assembling capsids in the nuclear compartment or whether it is added at a later stage which is discussed in the introduction. Depending on the cellular system used one might not obtain usable (i.e. comprising VP1-2) nuclear C capsids for this assay. Other essential tegument members that could play a part during entry, such as pUL37, are dependent on VP1-2 and their addition likely coincides or follows VP1-2 recruitment. Recruitment of nuclear transport factors is assumed to require VP1-2 but whether other tegument proteins are involved in this process is unknown. However, to avoid revisiting the longstanding issue of VP1-2 recruitment and since it would be difficult say what incoming capsids should look like I decided against using this approach.

(ii) Isolating capsids on the ‘way in’

To isolate capsids on their way to the nucleus should be the most relevant approach as virions will go through the physiological process of uncoating, de-tegumentation and binding of transport proteins. SV40, another DNA virus, that requires access to the nuclear compartment, only 1% of an inoculum was found to be infectious and much of it never reaches the nucleus (Flint 2009). I would define this 1% of the total capsids as the ‘good material’. Assuming a similarly poor scenario for HSV-1 (Frenkel et al. 1975) it is conceivable that, to obtain analysable material, cells probably need to be infected at very high MOI (50-100) followed by lysis. Much of that inoculum would never reach the

- 155 -

Biochemical analysis of VP1-2 binding partners nucleus and I would define that as ‘bad material’. For this experiment to be successful the ratio of ‘good’ to ‘bad’ material needs to high enough that MS based techniques or Western Blot can pick up the difference between wild-type and VP1-2ΔNLS capsids. There are probably two options where ‘good material’ could be isolated; either during their way through the cytoplasm but prior to docking at the nuclear pore or once bound at the nuclear membrane. A major factor complicating the first option is that the time and place where capsids pick up nuclear transport factors and how many do so eludes the scientific community. Recent data suggested that this might happen late during entry, potentially immediately prior to nuclear docking, since defective VP1-2NLS capsids can move along microtubules to accumulate in a juxta-nuclear position near the MTOC (Abaitua et al. 2012). This means the time frame to obtain ‘good material’ would be rather short. This could be a significant problem since the uncoating and migration towards the MTOC for the majority of capsids would have to be synchronous. To achieve some level of synchronicity infection can be done on ice. When the temperature is raised virions fuse more or less synchronously. However, this cooling procedure shatters the microtubule network.and re-assembly will almost certainly not be synchronous in every cell. Secondly, once the cytoskeleton is functional again some capsids will take substantially longer than others to navigate the cytosol and many will not get there. Thus it would be very difficult to obtain sufficient ‘good material’. Although importin-β localises to the cytosol, a very recent report found importin-β to form part of the NPC permeability filter within the central channel (Lowe et al. 2015). If capsids were to recruit importin-β at the nuclear pore, the time window for capsid isolation would be longer than in the just described scenario, since capsids were found to remain bound to nuclear pore for up to 6 hpi (Batterson et al. 1983). This would allow isolation of more uniform material (i.e. all capsids at the nucleus) while removing most of the ‘bad capsids’ that were not bound at the nucleus. However, although importin-β was shown to be required for nuclear docking (Ojala et al. 2000), it is unknown whether would be required for this prolonged attachment. I believe that, if the attachment is that strong, it would pose a considerable challenge to isolate these capsids.

(iii) Stripping virions and using them as ‘bait’

What can be done to improve the ratio ‘good’ to ‘bad’ capsids? I would argue that during natural infection a susceptible cell does not give up without putting up a fight. Capsids will be diverted into dead-end pathways or get degraded. Thus to increase the ‘good’ material I have to artificially bypass these cellular traps. To achieve this, one could envisage using whole capsids extracted from virion preparations. These can subsequently be used as an affinity matrix to recruit cellular binding proteins. Several groups have used a stripping approach using mild extraction buffers which allowed interactions with cellular proteins (Radtke et al. 2010; Wolfstein et al. 2006; Ojala et al. 2000). The methods

- 156 -

Biochemical analysis of VP1-2 binding partners involved multiple rounds of ultracentrifugation and resuspension steps and used sodium or potassium based buffers (depending on the assay). These were designed to mimic de-envelopment (detergent) and de-tegumentation (achieved through salt). The resulting capsids were subsequently shown to retain interaction with several proteins including microtubule motors. Because of the ease of obtaining sufficient ‘good’ capsids that retained interactions with proteins that are important during entry, I chose to use a similar approach. The challenge will be to balance the use of chemicals to achieve a level of de-tegumented capsids that would be proficient at interacting with the nuclear transport machinery. Herein, I present a simple one step protocol, termed SIEVH (Simultaneous Incubation and Extraction of Virus in Host) assay, to find capsid binding partners. For this approach to work the virion envelope has to be dissolved and most tegument should be removed. To achieve this two approaches were employed. (a) Purified virions were stripped using gentle extraction buffers (Radtke et al. 2010; Wolfstein et al. 2006) containing non-ionic detergents and different salt concentrations In this ‘uncoating’ scenario I explored the effect of ionic strengthon the removal of tegument proteins. The aim was to reduce the protein layer without removing the inner tegumentproteins VP1-2, pUL37 and pUS3, which are thought to remain attached during the early entry processes. (b) De-tegumentation is not thought to be a passive process which is caused by loss of the envelope but is contributed to by cellular factors and the reducing milieu. For example, many of the tegument proteins are targets of cellular kinases, proteases or connected by disulphide bridges. Thus I also explored the effects on capsid constitutionof cellular proteins. For this, cells were extracted in commonly used IP buffer but without ionic detergent (e.g. using RIPA containing only TX-100 but not SDS or deoxycholate).

5.2.2.2 Effect of salt concentration on capsid constitution

As elaborated above I decided to use the virion stripping approach. For this, concentrated virion preparations will be subjected to different extraction buffers for a set amount of time on ice. This should solubilise the envelope and some tegument proteins but leave the capsid intact. In turn, I attempted to sediment these capsids using a 45 min high speed centrifugation step. Only capsid-bound material should co-sediment during this procedure whereas solubilised proteins remain in the supernatant. The presence of representative capsid, tegument and envelope proteins in the different fractions will be tested by Western blot and silver stain. Standard extraction in PBS containing physiological salt concentrations and oxidising conditions showed near complete solubilisation of envelope protein gB and some other glycoproteins (Figure 28A/B, lane 1 and 2, buffer 1 and Figure 29, lane 1, blue arrowhead) after 15 min which was slightly improved with another 45 min of incubation (Figure 28B). As judged from the silver stain the extraction

- 157 -

Biochemical analysis of VP1-2 binding partners caused the release of a lot of virion proteins (Figure 29, lane 1, red arrowheads). The major capsid protein VP5 (Figure 29, lanes 1 and 2) was present only in the pellet giving rise to a strongly stained band around 170 kDa. The minor capsid component pUL25 (Figure 28A/B, lane 1/2), which anchors VP1-2 to the capsid, was also detected almost exclusively in the pellet fraction at both time points. This indicates that capsid integrity was undisturbed generally and also meant VP1-2 remained firmly attached to the capsid (Figure 28A/B and Figure 29, lanes 1 and 2). Most of its binding partner pUL37 was also found in the pellet at both time points (Figure 28A/B and Figure 29, lanes 1 and 2). Only a small percentage was found in the soluble fraction (Figure 28A/B, lane 1), which implies that at physiological salt concentrations both VP1-2 and slightly less so pUL37 remained firmly attached to capsids. The other VP1-2 binding protein VP16 appeared to be released to some extent as judged by the bands in lanes 1 and 2. However, the amount of VP16 per virion is much higher than that of pUL37 and the ratio between pelleted and solubilised VP16 was much smaller than the ratio of pelleted and solubilised pUL37, which suggests weaker binding of VP16 to VP1-2. The tegument proteins ICP4 and VP22 appeared to withstand isotonic extraction since both were recovered almost fully in the pellet fraction (Figure 28A/B and Figure 29, lanes 1 and 2). Again, time had little to no effect on the interaction of ICP4, VP16 or VP22 with capsids in buffer 1. In fact, time appeared to be at most a minor factor for solubilisation of outer tegument and thus could be limited to 15 min. Higher concentrations of NaCl (Figure 28A/B and Figure 29; lanes 3/4; buffer 3) may remove unwanted tegument proteins but could also be deleterious to capsid integrity since it would weaken ionic interactions. At first glance all proteins, except for gB, which was almost completely released in buffer 1 already, were more affected by increasing ionic strength of the extraction buffer as judged by Western Blot (Figure 28A/B). This was slightly more difficult to judge using silver stained SDSPA-gels, although there was evidence that at least three major bands that were abundant in the pellet fraction in buffer 1 were reduced or almost absent from the pellet recovered from buffer 3 (Figure 29; lanes 3/4, orange arrowheads). This effect was not increased using salt concentrations as high as 1M (Figure 30A, see buffer 6). While VP5 content (Figure 29, lane 4) was not affected by high salt concentrations the minor capsid protein pUL25 and slightly more so parts of the inner tegument, represented by pUL37 and ICP4, appeared affected by salt, although most of the three proteins still remained in the pellet. The slight release of pUL25 points to some disturbance of the capsid but this should be very limited since capsids were found to remain intact in similar extraction conditions (Wolfstein et al. 2006; Radtke et al. 2010) and only very harsh extraction conditions using guanidium hydrochloride were able to cause the release of viral DNA from capsids (Newcomb & Brown 1994). VP1-2 was also found to be released at a very low level compared to buffer 1 (Figure 28, see S3). This was probably not surprising since it binds to capsids via pUL25. The ratio of pelleted to solubilised bands for VP1-2 was also very

- 158 -

Biochemical analysis of VP1-2 binding partners similar to that found with pUL25. The effect on capsid and inner tegument proteins could not be detected by silver stain due to limited sensitivity of this technique. For the outer tegument protein VP16 the ratio of P3 to S3 was found to be increased compared to buffer 1 (Figure 28A/B and Figure 29, lanes 3 and 4, red/orange arrowhead at 58 kDa) indicating that this protein was released a bit more and might thus be interacting more weakly. In comparison pUL37, the other VP1-2 binding protein, was largely retained in the pellet. VP22 was similarly affected by ionic strength and I would say it was the most sensitive component tested in this assay since almost no release was observed in buffer 1. This places both VP22 and VP16 into the outer tegument category. To assess whether some proteins might just stick to precipitating capsids I attempted a higher stringency purification using different salt concentrations (Figure 30A). For this the procedure was repeated almost identically just instead of simply pelleting capsids I layered them onto a 35% sucrose cushion to get better separation of the soluble and pellet fractions. Only proteins whose interaction was strong enough to withstand the force of precipitation through the heavy sucrose would be sedimented with capsids. The soluble fractions clearly show that 500 mM and 1 M NaCl differed very little in their effects on solubilisation. The only noteworthy point was that recovery dropped significantly using the cushion whereas qualitatively there were no differences compared to the conventional sedimentation. In fact, this would only mean I would have to significantly increase the virion input to practically obtain the same result.

- 159 -

Biochemical analysis of VP1-2 binding partners

Figure 28 Effects of time, salt concentration, reducing milieu and cellular proteins on tegument and capsid components. Purified virions were extracted for (A) 15 or (B) 60 min using a PBS buffered solution containing 0.5% Triton X-100, protease inhibitors and 137mM NaCl as standard (buffer 1). The other buffers were additionally fortified as follows: Buffer 3 contains 500mM NaCl, buffer 4 contains 10mM DTT, buffer 5 contains approximately 1mg/ml of cellular protein. Lysates were subsequently centrifuged at 25,000 rpm for 45 minutes. Solubilisation and capsid integrity were tested by Western blotting with antibodies against representative proteins for the envelope (gB), outer tegument (VP22, VP16, ICP4), inner tegument (VP1-2, pUL37) and capsid (pUL25). S = soluble fraction; P = pellet after ultracentrifugation; number refers to buffer (e.g. P3 = capsid pellet of extraction with buffer 3). Approximately 10x more equivalent fractions were loaded for pellet compared to soluble.

- 160 -

Biochemical analysis of VP1-2 binding partners

5.2.2.3 Effect of reducing milieu on capsid consitution

A previous report suggested that the reducing potential of the cellular milieu contributes to removal of certain tegument proteins from the capsid. To mimic this effect I supplemented buffer 1 with 10 mM dithiothreitol (DTT), which effectively reduces disulphide bridges between proteins (Newcomb et al. 2012). This treatment (Figure 28A/B and Figure 29, lanes 5 and 6) had little effect on the composition of the pellet as judged by Western blot or silver stain. That is, the ratio of band intensities between the pelleted and solubilised fractions was comparable to that of buffer 1. In fact, it would have been very challenging to find any difference on a silver stained PA-gel. The only protein that could be modestly solubilised was VP22, which was only detected using Western Blot (Figure 28). This does not mean that other tegument proteins were not affected by DTT but it certainly suggests that ionic strength was better at removing outer tegument than reducing milieu alone.

5.2.2.4 Effect of cellular proteins on capsid constitution

Besides reducing power the cytoplasm contains cellular proteins which could through phosphorylation and dephosphorylation, limited proteolysis and/or competition for binding sites release some tegument from capsids. To account for this, I incubated virions in buffer 5 which contained all cellular proteins. Briefly, as in a simple IP, HEK293 cells were solubilised with an extraction buffer (here buffer 1). After 30 minutes of extraction, the cellular debris were removed and the supernatant recovered. This mixture of cellular proteins was designated buffer 5. When this cocktail was added to virions (Figure 28A/B and Figure 35, lanes 7/8), capsid integrity was undisturbed as judged by band intensity in the pellet fraction of the capsid components VP5 or pUL25. Recovery was comparable to buffer 1. Strikingly, VP1-2 and even more so pUL37 appear to be preferentially retained on the capsids as evidenced by Western Blot (Figure 28A/B). For pUL37 the band intensity in the soluble fraction (Figure 35A/B, lane 7) appeared much reduced. This might not have been visible for VP1-2 since the levels released in buffer 1 were already not detectable (lane 1). These data suggest stabilisation of the interaction between capsids and VP1-2 and pUL37. While VP16 appeared unaffected by cellular proteins, the VP22-capsid interaction was less stable (lanes 7/8) compared to extraction in buffer 1 which points to a role of cellular factors in selective removal of some outer and retention of inner tegument proteins. For this buffer, silver stain cannot yield interpretable information on removal or retention of most virion components due to the amount of cellular proteins present in the soluble fraction. However, the fact that the banding pattern of the pellet in lane 8 (Figure 29) did not change grossly indicated that capsids did not randomly co-precipitate cellular proteins. In fact only a few bands (green arrowheads, Figure 29) were selectively precipitated with capsids. These

- 161 -

Biochemical analysis of VP1-2 binding partners bands probably represent cellular binding proteins. This suggests that, using this method, it would be possible to identify cellular binding proteins.

Figure 29 Effects of different extraction conditions on the protein composition of capsids. Protein profiles were assessed with silver stained SDS-PA gels. Purified virions were extracted using the same buffers as described for Figure 28. Capsid components VP5, pUL25, pUL6, pUL17, VP19C and VP23, and selected tegument proteins (VP1-2, pUL37, ICP4) and glycoprotein gB are labelled. Blue (examples of solubilised glycoproteins) and red (examples of solubilised virion components) arrowheads demarcate protein bands that were released during the extraction procedure. Orange arrowheads (examples of solubilised outer tegument component) demarcate protein bands with altered band intensity compared to standard extraction (buffer 1). Green arrowheads demarcate additional bands, potentially of cellular origin, which were co-precipitated with capsids. Cell extract without virions (equivalent to buffer 5) and the resulting ultracentrifugation pellet (no virions) are also shown (last two lanes).

- 162 -

Biochemical analysis of VP1-2 binding partners

5.2.2.5 Effect of potassium on capsid composition

The above assays were looking at the standard co-IP conditions usually used to extract proteins or to look at protein-protein interactions. However, they still did not truly recapitulate cellular conditions subsequent to membrane fusion. Besides reducing milieu and proteins the cellular milieu also comprises potassium rather than sodium (Lodish et al. 2008). In fact some microtubule interaction studies specifically use potassium based buffers at a slightly lower pH of 6.8-7 (Radtke et al. 2010). To exclude potential extraction artefacts caused by the choice of buffer, the extraction procedure was repeated using a PIPES based buffer containing physiological [KCl] (137 mM) and a pH just below 7 (Figure 30B). A direct comparison of the pellets obtained by extraction-centrifugation in buffer 1 (PBS- NaCl) and buffer 7 (PIPES-KCl) (compare lanes 3 and 8) showed that the protein profiles looked comparable, although it appeared as though the KCl buffer increased the ratio of capsid components to major capsid protein VP5 (strong band at approximately 170 kDa) since the staining intensity appeared to increase in lane 7 (Figure 30) while VP5 was slightly reduced (compared to lane 3). In contrast, the amount of VP1-2 was reduced probably reflecting the lower amount of capsid protein VP5 recovered. Moreover, VP1-2 was probably also released during extraction since it was present in the input in lane 6 (reasonably strong band 2 mm from the top of gel). Addition of cellular proteins (lanes 5 and 10) to buffers 1/7 gave rise to similarly complex profiles with very few significant differences. Of note was that VP5 and VP1-2 appeared somewhat more abundant, which probably reflected a slightly increased amount of input virions. Interestingly, one band of likely cellular origin at approximately 65 kDa is reproducibly found in pellets extracted in both conditions (green arrowheads). This band (at approximately 65 kDa) was of particular interest since it falls into the approximate size bracket for α importins. Only one obviously altered band was found using buffer 8 (same a buffer 7 but containing cellular proteins) compared to buffer 5 (Figure 30B, lane 10, red arrowhead) which separated between 170-200 kDa. This means that in different extraction conditions different cellular proteins could bind to capsids underscoring the importance of salt types in this kind of assay. To get the most out of this technique both conditions likely have to be pursued in parallel.

- 163 -

Biochemical analysis of VP1-2 binding partners

Figure 30 Effect of high stringency extraction and purification, and potassium on virion extraction. (A) Virions were extracted with buffers 1, 3 and 6 (1M NaCl) as in Figure 29. The extract was either sedimented directly or through a 35% sucrose cushion which only allows dense protein ensembles to sediment. (B) Virions were extracted as before in standard buffer 1, buffer 5, buffer 7 (PIPES-KCl) and buffer 8 (PIPES-KCl containing cellular proteins). Red arrowhead demarcates an additional band of capsid bound protein likely of cellular origin which was only identified in

- 164 -

Biochemical analysis of VP1-2 binding partners buffer 8. The green arrowheads indicate the position of a reproducibly co-precipitated protein band of likely cellular origin. IN = input before extraction; S = soluble left after extraction and sedimentation; P = pellet 5.3 Discussion

Through work conducted for this chapter I was able to set up systems to identify interacting proteins of the VP1-2 N-terminus and of capsids using in vitro pulldown approaches. Using inducible cell lines I found several bands interacting with NT3 and/or NT3ΔNLS that were absent from the control. Although I did not identify any nuclear transport factors, that could be involved in capsid docking at the nucleus, an E3 ubiquitin ligase, DTX3L, could be confirmed as a VP1-2 interacting protein by MS and Western Blot. Interestingly, cells treated with the DNA damaging agent doxorubicin lacked 53BP1 recruitment to sites of DNA damage when NT3 and to a slightly lesser extent when NT3ΔNLS, were present. This points to a completely novel function of VP1-2 modulating the DDR. With a different approach I used a modified, one-step capsid extraction procedure to identify capsid interacting proteins. Distinct bands likely of cellular origin can be seen using silver staining which can now be analysed using MS. The data also showed that ionic strength and cellular proteins contribute to loss of certain tegument components during the extraction procedure, while at the same time cellular proteins were able to stabilise VP1-2 and pUL37 on the capsid.

5.3.1 Identification of VP1-2 interacting proteins using NT3 5.3.1.1 Technical considerations for the IP

While the pulldown of NT3 was a good technique to find interacting proteins of VP1-2 there were several issues using this system which could be improved using a few tweaks to the procedure.

(i) IP of NT3 worked poorly with very little recovery overall (ii) Importins were not identified using this technique (iii) Silver stained gels of the IP revealed many, potentially non-specifically precipitated protein bands

(i) One major problem with the IP approach was the obviously low recovery of bait protein from the cellular input (Figure 24A, lanes 8-10). I would argue that the antibody was just not good enough or was in some way sterically hindered by the position of the tag. The simplest way to solve this would be to switch antibodies to a commercially available rabbit polyclonal anti-V5 antibody or using a combination of the two. (ii) Since my initial aim was to identify nuclear transport receptors that bind the NLS I have reflected on the procedure. Was it even possible to identify nuclear transport factors with this pairwise approach? Can the assay be improved?

- 165 -

Biochemical analysis of VP1-2 binding partners

The assay was designed to reveal NLS interacting proteins through comparing one protein with a NLS to one that lacked this region using a Silac-MS approach. In retrospect, the two bait proteins cannot be used in a pairwise comparison since they localise to different compartments. A closer look at the localisation results show that NT3 was mostly nuclear (Figure 23B), possibly sequestered in complexes with nuclear proteins that were not compatible with importin binding. Also, importins are liberated from their cargo immediately after transport (i.e. only transient interaction). This implies that the proportion of NT3 actively shuttling into the nucleus at the time of lysis was probably minor and thus problematic to find. This was reflected by a large amount of differences seen in the band profile (Figure 24B, lanes 2-4) and MS results comparing NT3 and NT3ΔNLS. In fact, 99 and 32 proteins preferentially binding NT3 or NT3ΔNLS were identified, respectively. To solve this problem of localisation, NT3 and NT3ΔNLS could be expressed in bacteria followed by purification. Using these purified proteins as bait it would theoretically be possible to identify NLS binding proteins using heavy and light cell extracts. This system would compare both bait proteins like- for-like eliminating the confounding factor of localisation. Another group reported that another HEAT repeat containing protein, transportin-3, was only found to interact with its cargo if zwitterionic detergents were used. Thus the second tweak to this assay would be to include a suitable positive control (e.g. NT3 containing the SV40 lTag NLS) to check if it would at all be possible to precipitate any importins using the buffer conditions. Thirdly, to increase the proportion of NT3 that is actively shuttling I propose to insert a NES at the N- or C-terminus. This will allow nuclearly sequestered NT3 to re-enter the cytoplasm and to re- start the cycle of nuclear transport. Fourthly, the size of the bait could be reduced to below the nuclear diffusion limit. This would eliminate the need for a NES and additionally prevent other non-relevant parts of NT3 (e.g. USP) forming complexes that could preclude importin binding. If only the NLS region were fused to GST, for example, its small size would allow free travel between compartments. GST is generally expressed well and the pulldown very efficient using glutathione coated beads. This would also solve the recovery problems. Finally, to improve the chances of importin binding individual constructs encoding for tagged importins could be co-transfected. This could be done in combination with any of the previous suggested tweaks. Additionally, if the transfected importin-α sub-types could harbour deletions of their Importin-β binding domain this would enable them to bind cargo but render them unable to release it since this step requires importin-β. (iii) It would be much easier to identify binding proteins if there would not be as many contaminating and potentially non-specific interacting proteins co-precipitating with NT3 (Figure 24B).

- 166 -

Biochemical analysis of VP1-2 binding partners

This could be achieved by either introducing a second elutable or cleavable tag to the N- or C-terminus (e.g. His tag). Briefly, bait proteins could initially be purified using a cobalt resin followed by washes with increased concentrations of sodium chloride (up to 300 mM) to increase stringency and elution with imidazole or cleavage by a specific protease (depending on the cleavage site, e.g. precision protease). In a second round of purification a mixture of anti-V5 antibodies could be used to re- precipitate NT3 as before. Alternatively, I could use a similar approach with the unmodified NT3 construct since the cell lines already exist. For this, in a first round of purification, NT3 can be precipitated using anti-V5 antibody followed by elution with excess V5 peptide supplemented into the wash buffer. The eluate would then be re-precipitated using an anti-USP antibody.

5.3.1.2 NT3 interacting proteins

VP1-2 is a multifunctional protein comprising multiple functional domains. At least four functional domains can be found in its N-terminus alone which include a ubiquitin specific protease (Schlieker et al. 2005), a VP16 binding site (Mijatov et al. 2007), a NLS (Abaitua & O’Hare 2008) and another binding site for pUL37 (Mijatov et al. 2007). VP16 and pUL37 binding are critical for secondary envelopment (Desai et al. 2001; Fuchs, Granzow, et al. 2002) while the NLS enables capsid docking at the nucleus (Abaitua et al. 2012). The USP on the other hand is mostly dispensable for HSV-1 in cell culture (Bolstad et al. 2011) but was shown to fulfil accessory functions, including antagonising innate immunity (Wang et al. 2013), that probably contributes to successful lifelong infection of its host. As such inactivating mutations dramatically reduce the ability of PRV to establish latency (Böttcher et al. 2008). VP1-2 is a late protein and an N-terminal cleavage product was observed in infected cells, which can bind to ubiquitin (Kattenhorn et al. 2005; Schlieker et al. 2005). Since ubiquitination plays roles in proteasomal degradation it is conceivable that the USP facilitates stabilisation of target proteins in the wake of host protein shutoff. Ubiquitin, which can be linked into chains in multiple ways, fulfils additional signalling functions, usually to recruit other proteins to a particular ubiquitinated protein (reviewed in Bhoj & Chen 2009; Haglund & Dikic 2012; Jackson & Durocher 2013). Thus the USP could bind target proteins in order to modify their function. In my pulldown 199 proteins were identified that bound either NT3 or NT3ΔNLS or both but were not precipitated from a control cell line not expressing bait protein. Several interesting proteins, which cover functions in cell cycle control, transcription, mRNA metabolism, RNA binding, translation and DNA repair, could be identified with this screen. While HSV-1 might bring over 70 of its own proteins into the cell it certainly makes use of cellular factors to facilitate its replication (summarised in Tables 9-11). At least 4 proteins from the list of identified proteins were found to be required for HSV-1 infection, three seem to affect replication

- 167 -

Biochemical analysis of VP1-2 binding partners negatively, two are upregulated in at least one setting and at least one more binds directly or via previously confirmed interacting proteins. Although there is no evidence that VP1-2 (or here NT3) modulates mRNA metabolism or translation during HSV-1 infection it is well established that, as part of the viral programme, HSV-1 shuts off host protein production whilst maintaining viral protein synthesis (Roizman et al. 2013). Binding to proteins with annotated functions in cell cycle and DNA replication control which may contribute to the HSV-1 imposed cell cycle modulation (Flemington 2001) during infection to allow optimal replication of viral genetic material. An intriguing finding of the nuclear transport assay using GFP-R2 was that it localised to nuclear dots which looked like nucleoli (Figure 9A, panel III). Coincidentally, 32 proteins were found to bind nucleolar proteins. Upon closer inspection using a nucleolar localisation algorithm, the sequence around R1, specifically R5, resembles a nucleolar localisation signal. Since R5 was not required for nuclear localisation of NT6 (Figure 16B/C) but was essential for infection (Figure 18B) it would be intriguing to speculate that its nucleolar localisation, if a true phenomenon, might be one of the reasons why replication of HSV-1.VP1-2.ΔR5 was abolished. To test this, the rescue assay could be repeated with mutant NT6 comprising a different nucleolar localisation signal instead of R5 or in a different position in the background of ΔR5.

5.3.1.3 NT3 interacts with DTX3L

To validate that the MS did not erroneously identify certain binding partners I chose to follow up on DTX3L, an E3 ligase, that also bound to NT3ΔNLS (Table 9, Figure 27A, although weaker than NT3) and was separately verified in an excised band just below the NT3 band (Figure 24C). DTX3L was particularly intriguing since it and its binding partner PARP9 were found to be upregulated during HSV- 1 infection (Miyazaki et al. 2011) and through interferon signalling (Schoggins et al. 2011; Juszczynski et al. 2006) and DTX3L has both nuclear and cytoplasmic functions stimulating the DDR (Yan et al. 2009) and modulating the ESCRT pathway (Holleman & Marchese 2014), respectively. Both these pathways are likely required during HSV-1 infection. Since the anti-DTX3L antibody sample I used was not very good and the level of DTX3L in HEK293 very low (Figure 27C/F) it would be required to repeat this reciprocal pulldown with a better antibody, in other cells that express more DTX3L (e.g. Hela) or using a tagged DTX3L. It would also be beneficial to know where the interaction site lies. If it binds to a region after the USP it is more likely to be specific and, more interestingly, might be unique to α-HVs or even HSV-1. The interaction site can be mapped initially using sequentially shorter constructs of NT3. However, it has to be remembered here, that DTX3L might not bind directly to NT3. The MS data revealed that PARP9 was recovered at similar levels

- 168 -

Biochemical analysis of VP1-2 binding partners in the NT3 sample but to almost two-fold higher levels in the NT3ΔNLS sample. Since DTX3L and PARP9 interact (Yan et al. 2013) NT3 might precipitate it via PARP9. What could the binding of an E3 ligase mean for VP1-2 function? The most obvious reason for DTX3L to bind NT3 could be to ubiquitinate it. It is conceivable that in that way VP1-2 stability is affected since USP activity protects VP1-2 from degradation (Bolstad et al. 2011). Secondly, screens of the cellular ubiquitome (as retrieved from Uniprot and Genecards) identified at least 5 C-terminal ubiquitination sites in DTX3L, which can be auto-ubiquitinated to increase its own E3 ligase function (Takeyama et al. 2003). Thus, the USP might de-ubiquitinate DTX3L to reduce its activity. Thirdly and most intriguingly, the USP might affect targets of DTX3L by either preventing them to bind to the E3 ligase or by removing any attached ubiquitins immediately thereby inactivating particular pathways. This would not be a new mechanism since other cellular USPs can often be found in complex with an E3 ligase and quite often they reciprocally modulate each other’s activity or fine tune activity of ligase targets (reviewed in Ventii & Wilkinson 2008).

5.3.1.4 Proposed activity of VP1-2 in altering the cellular DDR

In addition to the pulldown studies that identified DTX3L (and PARP9) I was able to show that in the presence of NT3 (and slightly less so with NT3ΔNLS) 53BP1 foci formation in response to DNA damage was abolished. DTX3L and PARP9 are recruited to sites of DNA damage in PARP1 dependent manner and, once there, enable recruitment of 53BP1 through ubiquitination of Histone 4 at lysine 91 (Yan et al. 2009). As DNA damage signalling and the interferon response increase during HSV-1 infection, DTX3L and PARP9 are also upregulated (Taylor & Knipe 2004; Shirata et al. 2005; Mohni et al. 2010; Wilkinson & Weller 2004; Unterholzner et al. 2010; Juszczynski et al. 2006; Miyazaki et al. 2011) and thus potentially contribute anti-HSV-1 effector functions. The USP could modulate nuclear DTX3L/PARP9 function and consequently the nuclear environment in favour of the virus by altering DTX3L activity or by selectively de-ubiquitinating its target proteins. The following paragraphs will discuss the literature around DTX3L functions, the DDR in HSV-1 infected cells and implications arising from interaction of NT3 with DTX3L.

- 169 -

Biochemical analysis of VP1-2 binding partners

Figure 31 Simplified schematic of one of the pathways leading to 53BP1 activation and how HSV-1 could potentially modulate this part of the DDR. Adapted from Yan et al. (2013) with further information gathered from Grady et al. (2012); Lilley et al. (2011); Min et al. (2010); Jackson & Durocher (2013); Yan et al. (2009). (A) As soon as viral DNA synthesis increases, the amount of unusual DNA structures and DNA damage does, too, leading to PARP1 activation. (B) This leads to PARP9 dependent DTX3L recruitment, which ubiquitinates itself and histones (including H4K91 mono-ubiquitination). This induces structural changes in nucleosomes causing methylation at H3K79 and H4K20 to get exposed. (C) 53BP1 and BRCA1 are recruited by these histone marks. Immediate early HSV-1 protein ICP0 induces PARG degradation which deregulates PARP1 and consequently DTX3L activity. (D) The late protein VP1-2 (as the N-terminal product NT3) could remove ubiquitin from DTX3L or H4K91 (or others) which abolishes 53BP1 recruitment to nascent viral DNA. Reducing DTX3L auto-ubiquitination could reduce its activity with the same outcome.

- 170 -

Biochemical analysis of VP1-2 binding partners

There are multiple pathways that lead to 53BP1 recruitment to sites of DNA damage. The schematic in Figure 31 explains the relevant 53BP1 recruitment pathway (Yan et al. 2013) and how HSV-1 might subvert it. This involves the direct recognition of DNA damage or other unusual DNA structures (e.g. stable ssDNA and non-B form DNA) by PARP1/2, whose activity has also been implicated in the immune response (Galbis-Martínez et al. 2010) and apoptosis (Yu et al. 2002), followed by auto-ADP-ribosylation (reviewed in Bürkle & Virág 2013). This transient modification is quickly antagonised by PARG (poly-ADP-ribose glycohydrolase) activity which cleaves poly-ADP chains, either releasing chains of different lengths or monomers of ADP (Min et al. 2010). The PARP9-DTX3L complex accumulates at activated PARP1 foci transiently which leads to ubiquitination of histones including at H4K91 (Yan et al. 2009). As a direct result chromatin structure is opened revealing constitutive modifications including H3K79 (H3K79me) and H4K20 methylation (H4K20me). In turn 53BP1 and BRCA1 recognise the exposed modifications and accumulate to guide the decision of specific DDR pathways. 53BP1 supposedly promotes NHEJ while BRCA1 favours homologous recombination (HR) (Bunting et al. 2010; Nakamura et al. 2011). As soon as HSV-1 DNA enters the nucleus it faces the cellular machinery trying to silence expression of viral proteins and to set off a cascade culminating in the induction of an innate immune response. The DNA damage repair is not only pivotal to preserving cellular integrity during stress it is also part of the first line of defence against invading DNA. Intriguingly, HSV-1 inhibits NHEJ and homologous recombination (HR) but enhances ssDNA break repair (Schumacher et al. 2012). During HSV-1 infection DNA damage repair was found to be a two-edged sword with some part being required or beneficial while another was acutely anti-viral (Lilley et al. 2011; Taylor & Knipe 2004; Wilkinson & Weller 2004). Some parts of the DDR machinery are absolutely required for DNA replication contributing molecular functions such as helicase activity, dsDNA break and ssDNA protection and circularisation of genomes (Muylaert & Elias 2007; Lilley et al. 2005; Taylor & Knipe 2004; Wilkinson & Weller 2003; Placek & Berger 2010). Unsurprisingly, HSV-1 globally activates some DDR pathways to initiate and sustain an infection (Lilley et al. 2011; Taylor & Knipe 2004; Wilkinson & Weller 2004; Schumacher et al. 2012; Placek & Berger 2010; Mboko et al. 2012; Smith et al. 2014) For example it was shown that knock-down of γH2AX, the upstream signal for assembly of many DDR proteins, drastically reduced (Lilley et al. 2011) while ablation of NHEJ by deletion of Ku70 (Taylor & Knipe 2004) increased viral titres. DDR could cause inhibitory recombination events between the linear viral genomes which could be recognised as damage sites or could in fact enable replication (Smith et al. 2014; Muylaert & Elias 2007). Additionally, DNA damage and interferon responses are inherently linked (Mboko et al. 2012) which could allow an infected cell to induce an anti-viral state.

- 171 -

Biochemical analysis of VP1-2 binding partners

Interestingly, the DDR is facilitated by a network of post-translational modifications which includes many different ubiquitination events (Huen & Chen 2008; Thomson & Guerra-Rebollo 2010) which the VP1-2 USP could potentially modulate.

Figure 32 Simplified schematic of protein levels and PARP1 activity during infection. ICP0 is delivered into cells and transcribed very early on which leads to a steep rise in levels which plateau as immediate early transcription stops. ICP0 causes degradation of PARG and RNF8/168 proteins, which are involved in regulation of ADP-ribosylation and alternative 53BP1 recruitment, respectively. Since cellular protein synthesis is antagonised effectively their levels will not recover in the course of infection. With the onset of viral DNA synthesis DNA damage signalling is induced leading to PARP1 activation. VP1-2 levels only rise to very high levels after onset of DNA synthesis although a small amount is delivered into the cell with incoming virions. That means VP1-2 could modulate part of the DDR late on to maintain peak viral DNA synthesis. Information gathered from Min et al. (2010); Grady et al. (2012); Jackson & Durocher (2013); Roizman et al. (2013). Axes and levels not to scale.

ICP0 is highly expressed early on (Jones & Roizman 1979; Honess & Roizman 1974) (see Figure 29) and . During late infection the amount of viral DNA, DNA damage signalling and unusual DNA structures (e.g. concatemeric non-nucleosomal DNA) increase (Muylaert & Elias 2007; Schumacher et al. 2012; Grady et al. 2012). This leads to recruitment of PARP1 to replication compartments (Taylor & Knipe 2004) where it could activate the DDR through its ADP-ribosylating activity. ICP0 could inadvertently deregulate PARP1 activity through degradation of the regulatory protein PARG (Figure 31C) (Grady et al. 2012). In this scenario DTX3L could be recruited in a PARP1-dependent manner (potentially to replication compartments but this needs to be shown) and marks the DNA by ubiquitinating target proteins. With increasing amounts of viral DNA (>6hpi for example) and increasing PARP1/2 activity, this may make viral DNA more prone to DTX3L induced 53BP1 foci formation. 53BP1 in turn favours NHEJ which contributes to the anti-viral response to HSV-1 (Taylor & Knipe 2004). In this model the de-ubiquitinating activity contributed by the N-terminus of VP1-2, expressed after onset of DNA replication (Figure 32), could modulate DTX3L activity. This could occur via reversion of auto-

- 172 -

Biochemical analysis of VP1-2 binding partners ubiquitination by DTX3L or by de-ubiquitination of its targets. This theory is in agreement with the localisation of a NT3-like fragment, which, if transfected into cells prior to infection, showed strong signal corresponding to the NT3-like fragment in nuclear domains which are devoid of cellular DNA and could be replication compartments (Calistri et al. 2014). To test the hypothesis that VP1-2 might augment replication via modulation of DNA repair it first needs to be established which parts of the cellular DDR are affected and how VP1-2 might alter DTX3L function. To map the interaction site within VP1-2 co-precipitations and functional immunofluorescence (53BP1 foci formation) with smaller versions of N-terminal VP1-2 can be performed. To check if USP activity is required these studies should also be performed with USP- inactive N-terminal VP1-2 mutants. Additionally, the functional immunofluorescence analysis should be expanded to include various other DDR (including γH2AX, BRCA1, ATM, ATR, Ku70, etc.) which might reveal very specific (i.e. only 53BP1 affected) or more global modulation of the DDR. The second approach would look at HSV-infected cells comparing wild-type, USP-inactive mutants and ICP0-null virus. It would be worth introducing the USP-inactivating mutation (Bolstad et al. 2011) into a ICP0-null background, since ICP0 function is very dominant in cell culture systems. Since USP function in HSV-1 was previously shown dispensible for growth in cell culture it might require deletion of ICP0. Knocking out ICP0 makes the virus more susceptible to the effects of genome silencing and anti-viral factors (ICP0 functions reviewed in (Hagglund & Roizman 2004; Smith et al. 2011)) and reveal the potentially more subtle USP function. As DTX3L levels vary in different cells (Yan et al. 2009) and are increased by an immune response (Juszczynski et al. 2006) it will be worth including conditions such as interferon exposure and using cells with higher expression of DTX3L. For this experiment, I propose that DDR protein function in HSV-1 infected cells should be monitored using immunofluorescence. This can then be compared between the different virus mutants and correlated to the amount of virus released or viral DNA produced. The third approach, which should complement the other two, would look at cells lacking or overexpressing DTX3L and/or 53BP1. In these cells viral fitness can be evaluated using simple yield experiments or growth curves. Theoretically, overexpression of DTX3L might reduce viral fitness and viral DNA synthesis while knock-down should do the exact opposite.

5.3.2 Identification of cellular capsid binding proteins

With the preliminary work presented in this part I have shown that capsids that can be used for interaction studies can be obtained by simultaneous de-envelopment and de-tegumentation. Their tegumentation state, as judged by Western Blot against representative tegument proteins and silver stain profile, was influenced by salt concentration and cellular proteins but not by time of extraction,

- 173 -

Biochemical analysis of VP1-2 binding partners reducing agents more stringent purification. Additionally, using a one-step extraction-binding procedure, partially de-tegumented capsids, that bound proteins likely of cellular origin, can be obtained from small scale virion preparations. This technique might thus be useful to identify cellular capsid binding proteins. To optimise a protocol to obtain capsids that could be used to identify the nuclear transport receptors involved in capsid docking at the NPC one question should be addressed first; what do relevant capsids look like? Thus far the exact composition of capsids at the nuclear pore is unknown. But from immuno-EM and microtubule binding studies it can be inferred that these should ideally lose the outer tegument (including VP16 and VP22) but retain at least VP1-2 and possibly pUL37 as these proteins can be found on naturally de-tegumented capsids (Granzow et al. 2005; Radtke et al. 2010; Wolfstein et al. 2006). There could be several ways to identify factors that allow capsid nuclear docking and in my opinion I have chosen the approach that would yield the highest chance of obtaining the ‘right’ type of capsid to tackle this problem (explained in section 4.3.2.1). Briefly, nuclear capsid preparations are a good source of highly pure capsids but this requires a high amount of input material and purification (Radtke et al. 2010; Wolfstein et al. 2006). There is ongoing controversy where VP1-2 is added to the capsids (Radtke et al. 2010; Henaff et al. 2013) which may relate to sample handling or the cellular system of choice. This makes nuclear capsids a risky choice. Even more risky would be the isolation of capsids during natural infection since the exact time of recruitment of nuclear transport factors remains unknown. Additional problems with synchronicity of infection and the fact that many capsids never reach their destination make this endeavour a substantial challenge. In contrast, extraction of capsids in vitro circumvents many of these problems. For once all virions would contain the essential proteins VP1-2 and pUL37 and overall yield of the ‘right’ capsids will be much improved. However, this technique comes with its own challenges including stringency of extraction and multiple sequential purification steps. My extraction experiments showed that virions can easily be de-enveloped as indicated by loss of gB from the pellet fraction (Figure 34). In fact, it could be argued that envelope removal was slightly better than found in another study (Wolfstein et al. 2006). Moreover, capsid integrity was arguably undisturbed even at highest salt concentrations tested since almost all of VP5 (Figure 36A) and pUL25 (Figure 34) were present in the pellet and mostly absent from the supernatant. While time and a reducing agent had no effect on solubilisation of most tegument proteins, high concentrations of salt and less so cellular proteins were able to induce partial de-tegumentation which was detectable by Western Blot (VP22) for both conditions and by silver stain for salt concentration only. This was in agreement with previous reports on salt susceptibility of tegument proteins. From the silver stained PA-gels it could be concluded that recovery was much better using a simple spin rather than a sucrose

- 174 -

Biochemical analysis of VP1-2 binding partners cushion. Remarkably, the omission of stringent purification did not increase non-specific binding (retention) of viral proteins on capsids (Figure 36A). None of the extractions removed much of VP1-2 indicating that theoretically all these capsids should be able to engage nuclear transport factors relevant to NPC docking. However, it had been suggested that capsids extracted at salt concentrations below 500mM cannot engage microtubules (Radtke et al. 2010; Wolfstein et al. 2006) indicating steric hindrance or occlusion of binding sites by outer tegument. This in turn could indicate that the capsids obtained by extraction in buffer 1, 4 and, most importantly, 5 might not be useful for binding studies. Although these previous studies found microtubule interactions with capsids extracted in high salt (500 mM) they were not able to identify nuclear transport factors. A notable modification to the protocol of previous reports analysing capsid binding proteins was the use of simultaneous de-envelopment, de-tegumentation and incubation with cellular proteins combined with the omission of stringent purification steps (gradients or cushions). I would argue that the capsids I pelleted using my one-step approach were very similar to those found in a previous study (Wolfstein et al. 2006) as judged by Western Blot. The capsids obtained using the SIEVH protocol contained modestly more VP22 but similar levels of other tegument proteins. Most strikingly, VP1-2 and pUL37 retention was improved compared to standard extraction in the absence of proteins in agreement with the notion that cellular proteins stabilise certain capsid components (Henaff et al. 2013). To make more definitive statement it would be interesting to expand the study of the above described capsids. This would involve testing for the presence of more tegument proteins using immunoblots. Interestingly, cellular proteins co-sedimented with capsids extracted in cellular lysate. There were several distinct bands that were absent from pellets extracted in buffer 1 (Figure 36B). Slightly different results were obtained using potassium rather than sodium which could indicate that more physiological conditions might be superior. Whether the capsids obtained with the SIEVH assay are relevant is unknown. But what I have learnt was that these capsids could reproducibly bind cellular proteins which now could be identified using MS. The results from these studies will guide further optimisation of this system. Several aspects of the extraction could still be modified to potentially improve tegument removal. Increasing the temperature would be an obvious choice since on ice some tegument structures might be retained (low temperature slows/inhibits many processes including proteolytic degradation and modifications, and it lowers free energy which might prevent de-tegumentation). VP1-2 binding partner pUL37, for example, was released only at 37°C but not on ice in a previous study (Newcomb & Brown 2010; Henaff et al. 2013; Cardone et al. 2012). Although pUL37 is an inner tegument component it is not absolutely required for capsids to dock at the nucleus (Roberts et al. 2009), suggesting that increasing the temperature might improve the SIEVH protocol. However, increasing the temperature

- 175 -

Biochemical analysis of VP1-2 binding partners also increases degradation of proteins by cellular proteases. Thus the timing and concentrations of protease inhibitors might have to be adapted. Further studies should combine the tweaks of the extraction conditions with those mentioned for the NT3 pulldown approach (exogenous expression of importins including their mutants lacking the importin-β binding domain). The capsid approach is more suitable for Multiplex Silac-based MS analysis since wild-type and VP1-2ΔNLS capsids can be directly compared in their ability to bind cellular proteins. Briefly, mutant and wild-type capsids could be produced in cells labelled with two different isotopes (‘light’ and ‘heavy’). These capsids could be extracted in two different labelled cellular lysates (both labelled with different ‘intermediate’ amino acid isotopes). After the extraction and incubation both samples are mixed and purified as before. The pellets could subsequently be processed for MS analysis. This would yield a comparative binding profile which in theory should allow identification of NLS binding proteins.

- 176 -

General Discussion

6 Constructing a VP1-2 ΔNLS mutant expressing GFP-VP16 to study infected-cell fate 6.1 Background

Cells in a monolayer rarely act as isolated individuals but rather are interconnected via tight association between their cellular membranes which enable them to exchange factors and relay signals quickly thus allowing a monolayer to act in a concerted manner. Our group has previously shown that HaCaT cells (immortalised human keratinocytes) move as a group rather than single cells and uninfected cells are drawn into a developing plaque (Abaitua et al. 2013). To understand what is happening in the infected cells and in the adjacent non-infected cells it is vital to be able to follow them ‘live’ over time. The NLS virus is the ideal tool to study the adjacent (non-infected) cells as in this case these would not get infected. To allow for live cell studies of what happens in different cell lines, I created a VP16-GFP fluorescent derivative of the K.VP1-2.NLS virus. VP16 was chosen because it has been shown to function in the presence of a large tag, such as GFP, and it will be incorporated into nascent virions (Boissière et al. 2004; Heine et al. 1974; Newcomb et al. 2012). As it is expressed in high amounts late on it one can follow productively infected cells over time. To obtain a recombinant virus viral DNA and plasmid DNA (GFP-VP16) were co-transfected to allow homologous recombination to occur. Recombinant virus is subsequently serially purified on complementing cells until only GFP- positive plaques are formed.

6.2 Results 6.2.1 ΔNLS.GFP-VP16 replication is modestly affected by the presence of GFP

To confirm that I have obtained the correct recombinant virus I initially checked whether the fusion protein was expressed. For this, RSC and RSCUL36 cells were infected with high MOI of both the parental virus (VP1-2.ΔNLS) and one of the purified recombinant virus clones. Parental virus expressed a protein of 56kDa (predicted size) that could be detected with specific monoclonal antiserum (Figure 33A, lanes 8 and 11). This band was essentially absent from the recombinant virus (lanes 9 and 12). Expectedly, a band reacting with the same monoclonal antiserum was found to migrate at a position approximately 30kDa larger than VP16 indicating the presence of the GFP-VP16 fusion protein. To confirm this the same samples, separated on another SDS-PA gel, were blotted and the membrane probed with anti-GFP antibody (Figure 33A, lanes 1-6). The only observed band (lanes 3 and 6) had approximately the same molecular weight as that observed with anti-VP16 antibody (lanes 9 and 12).

- 177 -

General Discussion

Both mock infected cells and cells infected with parental virus lacked an anti-GFP reactive band (lanes 1, 2, 4 and 6). In conclusion, these data suggest that the correct virus was obtained. Adding a substantially sized tag to a protein usually interferes with function at some level. Additionally, it could be possible that the combination of VP1-2.NLS and VP16-GFP could affect virual fitness more than expected. Therefore, I initially looked at the ability of the recombinant virus to spread cell-to-cell which can be monitored by measuring plaque size (Figure 33C). All plaques were measured on RSCUL36 or HS30 cells as NLS and its derivative ΔNLS.GFP-VP16 do not form plaques on non-complementing cells. NLS mutant virus shows plaque formation was rescued by the complementing RSCUL36 line as plaques were of similar size as for NLS revertant virus (NLS.R). On complementing Vero cells (HS30) NLS virus also forms plaques, however, their size was modestly reduced compared to wild type (ΔNLS.R). This could be caused by different levels of complementing wild-type VP1-2 protein (here supplied in trans from the cellular genome) which competes with the mutant NLS form (encoded for by the virus genome) for capsid occupancy. It may require a certain critical ratio of wild-type VP1-2 to VP1-2NLS which might be lower in HS30 compared to RSCUL36 cells. Expectedly, NLS.GFP-VP16 plaques were less than half the size of ΔNLS.R and ΔNLS viruses on RSCUL36. On HS30 plaque size of the recombinant was still around half that of its parental virus but compared to wild type (ΔNLS.R) the effect was exacerbated with plaque size around 75% decreased. This indicated a fitness cost of GFP beyond the ΔNLS mutation (Figure 33C). To check whether the reduction in plaque size was due to the combination with the ΔNLS mutation I used two previously validated strain 17 recombinant viruses, comprising VP16 tagged at either the N- or the C-terminus, as controls (Figure 33D). This confirmed that adding a tag, such as GFP, to VP16 reduced plaque size up to two-fold in non-complementing cells. This defect was not specific to RSC cells which I used to create the recombinant virus, since a similar reduction of plaque size also occurred in Vero cells using both fluorescent VP16 fusion viruses (Figure 33D). Since these two wild- type viruses contain VP16 tagged at opposing termini it suggests that the position of the tag did not matter either. Another measure of viral fitness compares virus yield during one round of replication (single step growth curve). For this virtually all cells in a dish were infected (MOI of 5) and released virus titres determined over the course of 24 h. Comparing ΔNLS.GFP-VP16 to its parental VP1-2ΔNLS virus showed a reduction in yield of around one log unit (Figure 33B). Despite the large reduction in virus titre, reasonable amounts of infectious virions could still be recovered, which compared favourably with failure to obtain infectious virus from non-complementing cells. In conclusion, the modest reduction in plaque size and viral yield appeared to be a normal phenomenon when VP16 function as

- 178 -

General Discussion a transcription factor was impaired. For the purpose of creating viral stocks this modest impairment was of low significance since I was able to obtain viral stock with reasonably high titres.

Figure 33 Characterisation of ΔNLS.GFP-VP16 recombinant virus. (A) Western blot showing the presence of GFP-VP16, in cells infected at high MOI with the recombinant virus. Cells were lysed in Laemmli buffer and fusion protein was detected using anti-VP16 (right) or anti-GFP (left) (B) Single step growth kinetics comparing ΔNLS.GFP-VP16 and ΔNLS viruses. RSCUL36 cells were infected with high MOI (5) and total virus harvested at the indicated time points. Yield was titrated on RSCUL36 cells. (C) Plaque size comparison of wild type (ΔNLS.R = revertant), ΔNLS and ΔNLS.GFP-VP16 in two complementing cell lines, RSCUL36 and HS30. Plaque were stained with

- 179 -

General Discussion immunoperoxidase. Error bars represent standard deviation of 20-50 plaques. (D) Plaque size comparison to check how tagging VP16 N-terminally or C-terminally affects replication of another wild-type strain (17). Two non-complementing cell lines were infected at low MOI and plaques stained with immunoperoxidase. Error bars represent standard deviation from 20-50 plaques. 6.2.2 ΔNLS.GFP-VP16 virus does not spread in normal cells

To study the defect of the ΔNLS virus and the behaviour of infected cells I could now use the GFP-VP16 recombinant virus (Figure 34B). To illustrate the defect of the ΔNLS virus I have fixed and visualised a plaque assay of the ΔNLS parental and revertant viruses. For this complementing cells were infected at high MOI (>5) with virus that had been produced in complementing cells. Thus they carry a mixture of wild type and mutant VP1-2 on their capsids. This allowed all cells to become infected and consequently replicate the virus. This is illustrated in Figure 34A (first well of ΔNLS) where little crystal violet stain was retained by the fixed cells (i.e. they are dead). Upon further dilution of this virus stock the ΔNLS defect became apparent and plaques were only visible for the revertant ΔNLS.R (Figure 34A). When RSC cells were infected with ΔNLS virus it consistently looked like virus might be able to form small cell clusters that stained stronger with crystal violet (Figure 34D, white arrows).Thus, in order to find out what happens to infected cells and to see whether virus can actually spread to neighbouring cells, four non-complementing cell lines (HaCaT, RSC, RPE-1 and Vero) were infected with the recombinant virus (intermediate MOI of 0.2). The complemented virus (made in RSCUL36 cells) was able to infect cells (Figure 34B, white arrows) but then infection failed to progress in most non- complementing cells (see HaCaT, RPE-1 and Vero), whereas in the complementing cell line (Figure 34C) it could spread to form a plaque. Even after 72 h no progression was observed in HaCaT, RPE-1 and Vero cells even though this should have been sufficient time to form large plaques (compare to Figure 34C, 72 hpi). In fact, over time infected cells ‘rounded up’ but appeared to remain loosely attached, somewhat on top of the monolayer, for up to five days. This confirmed that through serial passage in complementing cells the ΔNLS mutation was still present in the recombinant ΔNLS.GFP-VP16 virus (Figure 34B). A closer look at the RSC cells, though, revealed that some virus spread might have occurred since several cells showed a positive signal for GFP (Figure 34B, RSC 48 h, 72 h and 120 h). These clusters increase in size very slowly, however, and neither the morphology nor the signal intensity of GFP resemble that of a typical plaque (Figure 34C).

- 180 -

General Discussion

Figure 34 Analysis of spread and plaque formation of ΔNLS.VP16-GFP recombiant virus over time. (A) RSC monolayers were infected with increasing dilutions of either ΔNLS or ΔNLS.R virus (from right to left decreased by 1 log unit per well) and the wells fixed and stained at 3 dpi with 1% formaldehyde and crystal violet. (B) Monolayers of non-complementing cell lines were grown to almost 90-100% confluence in 24 well plates and subsequently infected with 0.2 PFU/cell or 0.04 PFU/cell. Three fields with a reasonable amount of infection were followed for each cell line over five days and images of the same field were collected over 5 days. (C) A representative example of a ΔNLS.GFP-VP16 plaque on complementing RSCUL36 cells at 3 dpi. (D) Crystal violet stained RSC cells infected with ΔNLS virus showing small intensely stained cell clusters (white arrows).

- 181 -

General Discussion

6.2.3 Using NLS.GFP-VP16 as a tool to study the infected cell fate

I have already shown that cells infected with the KΔNLS.VP16-GFP virus appear to remain part of the cellular monolayer up to 120 hpi but it was unclear whether these cells were viable and what brought about cluster formation. It is conceivable that the defect of the ΔNLS mutation might have attenuated the recombinant virus affecting to outcome of infection (e.g. no cell death). Following ΔNLS.GFP-VP16 virus could yield valuable information on how infection progresses in individual cells. Thus to follow up on the curious finding of cluster formation in RSC cells and to study in more detail the live dynamics of virus infection, RSC cells were infected with ΔNLS.VP16-GFP (ΔNLS.VP16- GFP) virus at an MOI of 0.001. Three individual fields were imaged in phase and FITC channels every 20 min from 24-78 hpi and the images were collated into short movies. Two types of infected cells could be observed. Type 1 cells (movie 1, see attached CD) were already rounded up by approximately 24 hpi, as would be expected from infected cells. These cells appeared to be dead or dying and clusters did not originate from these cells. Type 2 cells, however, were not rounded but appeared somewhat healthier as judged by their general ability to retain attachment to the substratum. In fact, they appeared to ‘spread’ fluorescence to surrounding cells by an unknown mechanism (movie 2 and 3, both parts). As judged from the still images taken between 3-5 dpi (Figure 34B, RSC) the plaques appeared unconventional, with some uninfected (GFP-negative) cells in the centre of the cluster. The observed spread could have been casued by cell division, although, if you follow the time course frame- by-frame it revealed that the infected cell did not undergo mitosis. In movie 2 (part 1) the infected cell appeared to contain large vacuoles and a part of this cell containing some of these vacuoles had become separated from the main body of the cell. Intriguingly, parts of the infected cells seemed to have fused with neighbouring cells or parts of cells were taken up by uninfected cells through processes possibly involving endocytosis or macropinocytosis (movie 2, part 2 and movie 3, parts 1 and 2).

- 182 -

General Discussion

6.3 Discussion

The work in this chapter lead to the generation of a derivative of the VP1-2ΔNLS virus that expresses a fluorescently tagged tegument protein, GFP-VP16. This was achieved through recombination of exogenous DNA with the viral genome in infected cells. The recombinant virus expressed a protein approximately 30 kDa larger than the parental ΔNLS virus which could be detected with GFP and VP16 specific antisera. Adding a large tag to VP16 modestly decreased viral fitness of the recombinant and of two other wild type strains indicating the function of VP16 was modestly affected by the presence GFP. Cells infected with GFP-VP16 virus could be followed in cell culture for at least five days. Interestingly, the defective recombinant virus appeared to retain some limited infectivity in non-complementing RSC cells forming GFP positive clusters. Time-lapse fluorescence microscopy revealed an intriguing mechanism of ‘spread’ of viral components to neighbouring cells which could not be explained by infection. Using this recombinant virus it is now possible to investigate this phenomenon of cluster formation.

6.3.1 Viral fitness was impaired by GFP-VP16

I have shown through two independent measurements, single step growth kinetics and plaque formation, that the N-terminal fusion of GFP to VP16 affected viral fitness of the ΔNLS recombinant virus leading to a 65% and 74% reduction in plaque size on RSCUL36 and HS30 cells, respectively, and a ten-fold drop in virus production from RSCUL36 cells compared to parental virus (ΔNLS). This was surprising since the same fusion protein in wild type strain 17 had no effect on single step growth kinetics (Boissière et al. 2004). However, I could show that plaque size of two different strain 17 recombinants, expressing a N- or C-terminally tagged VP16, was also around 35% reduced in both cell types (Figure 34D). Since the presence of the ΔNLS mutation reduced plaque size of the parental virus in HS30 cells by approximately 50% already this indicated that the observed reduction of fitness was likely due to additive effects of expressing mutant VP1-2 and of the GFP-VP16 fusion protein. VP16 is an α-HV specific, essential transcription factor that is required to initiate viral transcription early on. It is delivered as part of the inoculum and translocates into the nucleus by virtue of its interaction with its transcriptional co-activator HCF (host cell factor) (La Boissière et al. 1999; Weinheimer et al. 1992; Campbell et al. 1984; Post et al. 1981). Once in the nucleus this complex assembles onto DNA via Oct-1 binding and recruits the transcription machinery via its extreme C-terminus (Milbradt et al. 2011). Viral fitness could have been reduced due to disruption of these VP16 interactions. However, since this was a N-terminal GFP fusion and the Oct-1/HCF binding regions

- 183 -

General Discussion were mapped to the C-terminus on the other side of the protein structure (Liu et al. 1999) this implies that this function was likely not affected. VP16 also contains regions that allow it to be packaged into virions, which lie closer to the N- terminus (Ace et al. 1988). Thus, it is conceivable that the introduction of an N-terminal tag like GFP could have affected virion incorporation. However, this did not seem to be the case for the recombinant wild-type (strain 17) viruses also tested (Figure 34D) as part of this chapter (Boissière et al. 2004). Although the available data give no indication of why VP16 function might be affected, one has to take into account the formation of a considerable complex around the transcription initiation site (Milbradt et al. 2011). It is conceivable that GFP causes steric hindrance that might affect this process and thus transcription initiation causing the observed reduction in viral fitness.

6.3.2 Cluster formation

As illustrated in Figure 35B (and with arrows in Figure 35D) ΔNLS viruses lead to formation of clusters of GFP positive cells. This phenomenon was also previously detected with indirect immunofluorescence using an anti-gC antiserum in rabbit cells infected with a PRV mutant virus lacking VP1-2 (Fuchs et al. 2004). The authors concluded that this was not due to a division event but rather production of some infectious virus particles. It is noteworthy that I only observed clusters consistently in rabbit skin cells which suggests that this might be a rabbit cell specific phenomenon and potentially dependent on cellular processes either of the infected cell or of neighbouring uninfected cells. The movies I took were in agreement with the conclusions drawn by Fuchs et al. (2004) that cluster formation was not brought about by cellular division. However, it did not conclusively resolve whether these cells actually got infected and were producing virions. Although the question remains open without further experiments, four scenarios might explain this phenomenon. (i) Infected cells produced a low amount of infectious virions that spread through residual binding of nuclear transport factors by VP1-2ΔNLS or another capsid (or tegument) component, or via direct binding to the nuclear pore through capsid and/or tegument components (ii) Wild-type VP1-2 from incoming, complemented virions can be incorporated into nascent capsids to allow some infectivity (iii) ΔNLS capsids can find an alternative route into the nucleus (iv) Neighbouring cells take up infected-cell material actively

- 184 -

General Discussion

(i) Production of low amount of infectious virions?

The authors of the study conducted by Fuchs et al. (2004) suggested that a small amount of infectious virions could be produced that infect neighbouring cells. This was a reasonable conclusion because VP1-2ΔNLS might retain some importin binding activity. A closer look at the raw data of the MS results of NT3ΔNLS (see Appendix 5, importin-8) revealed that importins may potentially still bind the ΔNLS isoform but affinity and binding strength would be severely reduced in the absence of an NLS (Cutress et al. 2008; Hodel et al. 2006). Thus, there could be the possibility that even capsids carrying exclusively mutant VP1-2.ΔNLS might be able to use nuclear transport factors to get to the nucleus at highly reduced frequency. Alternatively, there are several candidate proteins that might lend their functions to allow docking of ΔNLS capsids at the NPC at very low frequency (see Table 11). These proteins might be part of the capsid or the tegument. Anderson et al. (2014), for example, showed that nuclear C capsids, which in their setting did not comprise VP1-2, bound nuclear pores almost as well as capsids extracted at 0.5M KCl. This implies capsids might bind to nuclear transport factors or the NPC in the absence of tegument proteins. The only capsid proteins that contain a recognised NLS are VP19C and VP24. The former is buried between capsomeres and the latter within the mature capsids, likely rendering them unable to bind any transport factors. Although not comprising functional NLSs, VP5 and VP26 cover the whole capsid surface and thus might be able to recruit ‘adapters’ that in turn allow nuclear docking. However, VP5 is also likely buried and deletion of VP26 was tolerated well in cell culture (Desai et al. 1998) making both unlikely candidates. The Portal pUL6 and vertex associated protein pUL25, despite lack of a confirmed NLS, are nuclear proteins and according to one report can bind nucleoporins hCG1 and Nup214, respectively (Pasdeloup et al. 2009; Rode et al. 2011; Copeland et al. 2009). This interaction might occur even in the absence of the VP1-2 NLS or full length VP1-2 to enable NPC docking. Other likely proteins to recruit nuclear transport receptors belong to the tegument. Of all tegument proteins only three do not localise to the nucleus (products of UL11, UL46 and US2) and one only translocates to the nucleus during late on during the cell cycle (VP22). All others can access the nucleus (Table 11), although only seven are known to contain functional NLSs which could directly recruit nuclear transport factors (see Table 11 including references). The most likely candidates are ICP0, pUS3, pUL14, pUL16, pUL21, pUL37 and VP1-2 which form part of the inner tegument shell, although all with the exception of VP1-2 and pUL37 were found to be dispensable for infection in cell culture (Desai 2000; Desai et al. 2001). Although VP1-2 is thought to be the assembly platform for tegument proteins, there is some evidence that some inner tegument proteins bind to capsids in the absence VP1-2. For example, pUS3, ICP0, ICP4 and pUL14 were found on nuclear capsids amidst

- 185 -

General Discussion controversy over when VP1-2 is recruited (Radtke et al. 2010). The capsid interaction of said proteins was strongest for pUS3, pUL14 and ICP0 while pUL16 and pUL21 could be removed from capsids more easily with higher salt concentrations. Intriguingly, pUL14 had previously been implicated in nuclear import of capsids (Yamauchi et al. 2008). The authors report that fewer UL14-null capsids bound to nuclei of U2OS cells. VP1-2 is a prerequisite to bind pUL37 (Klupp et al. 2002). Although pUL37 appeared to be essential for HSV-1 replication (Desai et al. 2001) it was not required for the closely related PRV (Leege, Granzow, et al. 2009). Since pUL37 also does not contain a confirmed NLS it is an unlikely candidate for rescuing the ΔNLS phenotype. Since cell clusters were observed with PRV which lacks VP1-2 (Fuchs et al. 2004) and VP1-2 likely forms an assembly platform for outer tegument proteins (Bucks et al. 2007) this excludes a function of most outer tegument proteins during nuclear docking. Of the other tegument proteins the only realistic candidates are ICP4 and ICP34.5. ICP4 may already be recruited to capsids in the nucleus and the ICP4-capsid interaction appeared stronger than that of the outer tegument protein VP22 (see 5.2.2, buffer 3 extraction of capsids), as salt and reducing environments only allowed partial solubilisation of this protein (also in Radtke et al. 2010; Wolfstein et al. 2006). Being an essential transcription factor, it contains a functional NLS and to improve its own transport it might hijack the capsid as a taxi to the nucleus and thus might also allow capsid docking at the NPC. ICP34.5, a neurovirulence factor, also comprises a functional NLS but reports are scarce on when ICP34.5 is recruited to capsids. The capsid interaction tolerated high salt treatment reasonably well (Radtke et al. 2010) and could mean it is localised closer to the capsid. Whether it remains attached during transit has not been shown but it appears likely that tegument removal is not 100% complete on all capsids and thus it might allow capsid docking at the nucleus in the absence of VP1-2.

- 186 -

General Discussion

Table 11 Localisation and presence of NLSs in documented capsid and tegument proteins. Green script indicates proteins that were not required for retrograde transport (Antinone et al. 2006) and red boxes highlights proteins that show no experimentally confirmed nuclear translocation or are thought not to contain an NLS. A systematic review of the ability of most HSV-1 proteins to translocate EYFP to the nucleus as part of a fusion protein was undertaken by Xing et al. (2011).

ORF Protein name NLS Nuclear localisation observed? References

UL7 ? yes (Xing et al. 2011; Roller & Fetters 2015) UL11 no no; nuclear membrane (Xing et al. 2011) UL13 ? no/yes (Xing et al. 2011) (Xing et al. 2011; Ohta et al. 2011; UL14 yes yes Yamauchi et al. 2002) UL16 ? yes (Xing et al. 2011) UL21 ? yes but mostly cytoplasmic (Xing et al. 2011) UL23 ? yes (Xing et al. 2011) (Xing et al. 2011; D. S. McNabb & UL36 VP1-2 yes yes but mostly cytoplasmic Courtney 1992) UL37 ? yes but mostly exported (Xing et al. 2011; Watanabe et al. 2000) UL41 Vhs ? yes (shuttling) (Shu et al. 2013)

UL46 VP11-12 ? no (Xing et al. 2011; Kopp et al. 2002) (Donnelly & Elliott 2001; Xing et al. 2011; UL47 VP13-14 yes yes Kato et al. 2011) Tegument (Yedowitz et al. 2005; Xing et al. 2011; UL48 VP16 yes yes Salsman et al. 2008; Morrison et al. 1998) (Elliott & O’Hare 2000; Xing et al. 2011; UL49 VP22 ? yes (during M phase) Blouin & Blaho 2001) UL50 ? yes (Xing et al. 2011) UL51 ? yes (Roller & Fetters 2015; Xing et al. 2011) UL55 ? yes (Yamada et al. 1998; Xing et al. 2011) US2 ? no (Xing et al. 2011) US3 ? yes (Reynolds et al. 2002; Xing et al. 2011) (Salsman et al. 2008; Cheng et al. 2002; RL1 ICP34.5 yes yes Harland et al. 2003) RL2 ICP0 yes yes (Mullen et al. 1994) RS1 ICP4 yes yes (Mullen et al. 1994) UL19 VP5, ICP5 no yes but mostly cytoplasmic (Rixon et al. 1996; Salsman et al. 2008) (Rixon et al. 1996; Salsman et al. 2008; Li UL38 VP19C yes yes et al. 2012; Adamson et al. 2006) UL18 VP23 no yes (Rixon et al. 1996; Salsman et al. 2008)

UL26 VP24 yes yes (Salsman et al. 2008) Capsid (Yamauchi et al. 2001; Salsman et al. UL35 VP26 no yes (via pUL14) 2008) (Burch & Weller 2004; Nellissery et al. UL6 Portal ? yes 2007) UL25 ? yes (Cockrell et al. 2011)

(ii) Complementation in trans by wild type input VP1-2?

Our group has already shown that when non-complementing cells (e.g. RSC or Vero) were infected with a complemented ΔNLS virus they become infected and produce virions normally. The

- 187 -

General Discussion resulting non-complemented virions in turn showed approximately 1000 fold lower infectivity (Abaitua et al. 2012). When cells are infected with complemented virus these carry some wild-type VP1-2 on their capsids. One could argue that some residual wild-type VP1-2, derived from the initial inoculum, might be integrated at a low frequency into nascent capsids and thus allow for very low infectivity and limited cluster formation. However, it would be unlikely for this to be the case. Firstly, the infection was undertaken at low enough MOI to limit complemented input capsid to a few capsids per infected cell. Secondly, VP1-2 is not the most stable protein in the absence of infection and shows breakdown and turnover (Bolstad et al. 2011) in cells that are not yet at the peak of replication (D. S. McNabb & Courtney 1992). Thirdly, since in complementing RSCUL36 cells ΔNLS viral titres were still around 5 fold reduced (80% less) compared to a wild-type virus (Abaitua et al. 2012) it argues that the two VP1-2 isoforms compete for capsid binding and implies a negative, if not partially dominant-negative, effect of mutant VP1-2. The number of VP1-2 binding sites per capsid vertex is estimated to be 5, which corresponds to the number of UL25 copies (Newcomb et al. 2006; Coller et al. 2007). Thus up to 60 copies of VP1-2 could bind to all 12 vertices. It might require more than one copy, or even the majority (or a maximum of 5 copies), of wild-type VP1-2 per vertex to allow for efficient nuclear transport factor binding. Assuming the presence of equal amounts of mutant and wild-type VP1-2 and that their likelihood of capsid binding is exactly the same, the chance of having at least one wild-type VP1-2 at any one vertex is 96.9% but drops to 68% if you want at least one wild-type copy at each vertex. This means 1 in 10 capsids would carry at least one wild-type VP1-2 per vertex. Thus it appears likely that more than one wild-type VP1-2 per vertex would be necessary to provide the necessary nuclear transport function. In the non-complementing scenario, when VP1-2 complemented ΔNLS virions infect non-complementing cells, a large amount of mutant VP1-2 (generated during infection) and a low amount of wild-type VP1-2 (from the input) compete for capsid binding. This would lower the chance to have a wild-type VP1-2 at the vertex. Taken together there is only a low likelihood that residual input VP1-2 would allow sufficient infectivity to cause the observed cluster formation.

(iii) An alternative nuclear entry pathway?

Interestingly, I found it difficult to obtain chimeric VP1-2 viruses (containing homologous R1s) in Vero cells but succeeded in RSC cells. Rabbit cells might harbour specific properties to allow VP1-2NLS mutants to propagate. Interestingly, it has previously been shown that in rabbit cells an otherwise highly debilitated PRV mutant, which lacks the capacity to form the essential nuclear egress complex, made up of pUL31 and pUL34, and consequently was unable to exit the nucleus, can utilise an alternative route to overcome its defect (Grimm et al. 2012). Passaging this virus gave rise to a revertant that locally dissolved the nuclear membrane (during infection) to reach the site of secondary

- 188 -

General Discussion envelopment (Table 3, introduction). It might be possible for HSV-1 ΔNLS capsids or other delivered proteins (as part of the inoculum) to locally disrupt the integrity of the nuclear membrane. Through these perforations capsids or released DNA could directly enter uninfected nuclei. On the one hand, DNA release required the concerted actions of importin-β and the RanGTP cycle in addition to the ability to dock at the NPC, implying recruitment of nuclear transport machinery must be required for this process. On the other hand, Horan et al. (2013) has shown that HSV-1 capsids can be degraded by the proteasome which causes a release of HSV-1 DNA. In this hypothetical scenario cytoplasmic or nuclear proteasomes might act to release infectious viral DNA which can enter through nuclear pores or perforated NE. Alternatively, the virus could exploit NEBD during division. These scenarios are entirely hypothetical and would require further investigations relating to infectivity of non-complemented stocks in RSC cells and whether these could be serially passaged. If during serial passage a spontaneous revertant arose which retained the ΔNLS mutation then one could find out which alternative pathways may be utilised. Such a virus lends itself to ultrastructural investigation using EM which might indicate any of the above mentioned phenomena (e.g. capsid degradation, NEBD, nuclear capsids) Why rabbit cells appear to be more conducive to VP1-2 mutant viruses is unknown and could also warrant further investigation (e.g. different nucleoporins, nuclear transport factor types and specificities or different capsid de-tegumentation kinetics).

(iv) Active uptake of infected-cell material by uninfected cells?

HSV-1 infected cells are generally dead within one day when infected with wild type viruses. Using a mutant like ΔNLS.GFP-VP16 might slow down replication but cells still looked like they underwent terminal processes as seen in movies 1-3. When cells undergo apoptosis their remnants, exposing ‘uptake signals’ including phosphatidylserine (Fadok et al. 2001), generally get taken up by other cells which is not solely an immune cell function. Human keratinocytes, for example, were observed to internalise microspheres up to a diameter of 20m in vitro (Morhenn et al. 2002). Thus, I propose, that these filmed uptake or fusion events could either be a normal clearance and recycling process of cells in a monolayer which may look like spread of infection between neighbouring cells. One could speculate that this may be a way HSV-1 could spread from cell to cell.

- 189 -

General Discussion

6.3.3 Proposed experiments to solve the question of cluster formation

To test if ΔNLS virus (or even ΔVP1-2) deposits DNA into neighbouring RSC cell nuclei several complementary techniques could be used. Firstly, it has to be established whether ΔVP1-2 virus can deposit capsids into neighbouring RSC cells by EM. Their localisation would be of particular interest. Are they near the nucleus and, if so, can any be seen docking at the NPC? Following this, it could be checked if any viral DNA can be found in uninfected nuclei of a cluster by fluorescence in situ hybridisation. This will give a definitive answer as to whether infection could be the cause of the observed cluster formation. To exclude phagocytosis or macropinocytosis, cells could be tested for the presence of relevant receptors (e.g. for phosphatidyserine). Also, specific inhibitors to signalling pathways could be employed to see if this process can be blocked. Staining for certain endosomal markers for analysis by immunofluorescence or EM could ascertain whether these pathways were involved. The arguably most interesting experiment would be to obtain a revertant virus through serial passage on rabbit cells. This might reveal previously unknown functions of HSV-1 proteins or even a new way of HSV-1 to spread from cell-to-cell

- 190 -

General Discussion

191

General Discussion

7 General discussion 7.1 Summary

For HVs to establish a life‐long infection their genome needs to reach the nucleus of a susceptible cell. The last step of entry involves binding of nuclear transport factors which are required to dock at the nuclear pore and subsequently uncoat the genome. I set out to elucidate the mechanism of nuclear docking of HV capsids. I met the original objectives of

(i) defining a suitable system to screen for NLSs in VP1-2 homologues (ii) identifying the relevant NLS motifs in homologues of all HV sub‐families both by in silico analyses and through empirical testing in two different in vitro systems (iii) showing NLS functionality of VP1-2 homologues in the context of HSV‐1 replication (iv) identifying functional determinants in the HSV‐1 NLS region required for nuclear localisation (v) characterising functional determinants of the HSV-1 NLS required for viral replication (vi) isolating and characterising recombinant viruses harbouring mutant and chimeric VP1-2 NLS regions (vii) optimising a simple one‐step protocol to isolate viral capsids in complex with cellular interacting proteins which can subsequently be analysed by MS (viii) using mass spectrometric analysis to identify cellular interacting proteins of VP1-2 (ix) confirming one VP1-2 interacting protein by IP and a functional DDR assay (x) creating a fluorescently tagged VP1-2.ΔNLS recombinant virus to study infected-cell behaviour

192

General discussion

7.1.1 Nuclear localisation and viral replication

The data in the preceding chapters has shown that VP1‐2 homologues of all herpesvirus sub‐ families contain a motif at an analogous position, which could be predicted using a variety of in silico sequence analysis tools. These putative NLSs function in nuclear transport assays in the context of a heterologous protein and in a long N-terminal fragment of VP1‐2 itself (see summary in Table 12).From the structural organisation of R1 (or R8) it can be assumed that, with the exception of HHV-8.R8, all R1 equivalents identified in the α‐, β‐ and γ1 HV sub‐families comprise a bpNLS, which could functionally substitute, albeit only partially, for the equivalent HSV‐1 region. More importantly, for these bpNLS motifs (VZV, HCMV and EBV R1s) purification of recombinant viruses was possible. Although they all exhibited a generally modest reduction in viral fitness (reduction in plaque size, comet tails and yield of infectious virions) on several cell lines, their replication was significantly impaired in Vero cells compared to wild type virus. The defect of these viruses could be precipitated by a combination of delayed entry and replication kinetics and reduced production of infectious virions.

Table 12 Summary of the functionality of the homologous NLS regions. Efficiency is indicated compared to wild type virus (+++ same as wild type, - demarcates negative result).

Nuclear localisation Function during HSV replication Type of Virus Region NLS ΔNLS Virus Viral β-gal VP1-21-1875 Rescue prep replication

HSV-1 R4 +++ +++ mpNLS - (ΔR5) - -

mpNLS, HSV-1 R1 +++ +++ +++ Yes +++ bpNLS

VZV R4 - Not done Not done Not done Not done

VZV R1 +++ ++ bpNLS ++ Yes +

HCMV R1 +++ +++ bpNLS + Yes +

EBV R1 +++ +++ bpNLS + Yes +

HHV-8 R1 - - - -

HHV-8 R8 +++ +++ mpNLS - -

R1.Myo 2x mpNLS, HSV-1 +++ +++ + Yes + NLS bpNLS

- 193 -

General discussion

My initial hypothesis that NLS function would be the sole requirement for rescue of the ΔNLS phenotype and thus viral replication had to be modified; NLS function was absolutely required but not sufficient for rescue. There was correlation, however, with the organisation of R1 and ΔNLS rescue. All motifs that contained a bpNLS, that is, two basic patches separated by a variable spacer, were able to function in the context of HSV-1 replication.

Table 13 Summary of the functional determinants of the HSV-1 NLS region. Efficiency is indicated compared to wild type virus. (+++ same as wild type, - demarcates negative result)

Function during HSV Nuclear localiation replication Region Mutation NLS β-gal VP1-2 Virus prep 1-1875 Rescue R1 +++ +++ +++ Wild type

R4 +++ + (R5) - (ΔR5) -

R5 - - (ΔNLS) - (ΔNLS) - R4 K428R - Not done Not done - R1 K428R +++ + ++ Yes R1 K428A - - - - R1 ΔNLS - - - - R1 ΔR5 +++ + - - R1 mutR5 +++ + - -

For the HSV‐1 motif (see Table 13) I established that the R1 region contains a mpNLS which is part of a larger bpNLS structure. Although both NLS types are sufficient to translocate a heterologous protein into the nucleus, integrity of R1 was required for infectivity. In the context of VP1-2 and during replication R4 (the first basic patch) contributes the majority of NLS function, given that ΔR5 but not ΔR4 constructs could confer nuclear localisation. These data corroborate data on the PRV NLS which led to the conclusion that the HSV‐1 NLS, and probably that of all other α-HVs, likely acts as a bpNLS during infection.

7.1.2 Interaction studies

To identify interaction partners of the VP1‐2 NLS I established cell lines inducibly expressing nuclear and cytoplasmic (ΔNLS) versions of the N‐terminal quarter of VP1‐2, which can be used to identify VP1‐2 interacting proteins. Initial analysis identified VP1-2 interacting partners although these were probably not directly dependent on the NLS for interaction but likely on the localisation of the VP1‐2 fragment. This work is relevant to infection since an N‐terminal fragment of VP1‐2, comprising

- 194 -

General discussion the USP and probably the NLS, may be released during infection (Kattenhorn et al. 2005), which might fulfil important functions. For the purpose of this thesis I identified and validated the interaction of the VP1-2 N-terminus with DTX3L, an important regulator of the DDR, by IP. Additionally, I showed that NT3 and less so NT3ΔNLS, likely by virtue of their interaction with DTX3L, abolished 53BP1 recruitment in response to doxorubicin. This could have implications during infection since several aspects of the DDR appear to be modulated at several stages of the HSV‐1 life cycle (discussed earlier). Additionally, I optimised a modified approach to identify VP1‐2 NLS interacting proteins of relevance during the entry process by using a one‐step capsid purification protocol. I showed that extraction of capsids from purified virions reproducibly and selectively removed selected glycoproteins and outer tegument proteins. This was enhanced with salt concentrations above 500mM but not by reducing agents. Compared to extraction in standard lysis buffer, extracting capsids in the presence of soluble cellular proteins stabilised important capsid tegument interactions (VP1-2, pUL37) while destabilising others (VP22). Intriguingly, few distinct bands probably of cellular origin, can be identified using silver stain. Thus, this work forms the basis of a simple one-step protocol to isolate de‐enveloped and partially de‐tegumented capsids complexed with potentially relevant cellular proteins by co-extracting purified virions and intact cells. Although I was not able to identify the nuclear transport factors involved in HSV‐1 entry, this work led to the preliminary identification of a potentially novel and unexpected function of the structural protein VP1-2 in modulating the DDR. The described NT3 based expression system can be used in future to pursue interesting questions concerning VP1‐2 (USP) functions during HSV-1 infection.

7.2 Discussion 7.2.1 Conclusions

From the work undertaken for this thesis four reasonable conclusions can be drawn. (i) Taking into account the relative conservation of VP1-2 domain structure, the conservation of the basic patches and the localisation results it is very likely that all HVs contain an NLS within the linker between the USP and the central domain. (ii) Since the VP1-2 sequences from all homologues except for HHV-8 were able to rescue the ΔNLS defect, it appears that this motif likely functions to guide incoming HV nucleocapsids to the nuclear pore. (iii) NLS strength did not correlate with ΔNLS rescue and viral fitness of the recombinant viruses if foreign sequences were inserted into VP1-2. However, if the wild type NLS was weakened through mutation viral replication was reduced.

- 195 -

General discussion

(iv) Since the R5 region was essential for viral replication but was not absolutely required for nuclear localisation it could indicate that additional functions might be attributed to this region.

7.2.2 Model of nuclear docking

In a simplified model of HV nuclear docking (exemplified by HSV-1) HV capsids are deposited into the cytosol and tegument dissociates except for VP1-2 and a few other proteins. These capsid structures engage the cytoskeleton and exploit cellular motor proteins such as dynein to move towards the MTOC in retrograde fashion. To get to the nucleus capsids might have to switch their preferred transport direction (anterograde) to move towards the nucleus. Since ΔNLS capsids accumulate at the MTOC there are two not mutually exclusive possibilities where capsids could engage the nuclear transport machinery. They could pick up importins during or immediately prior to transport from the MTOC to the nucleus which would mean ΔNLS capsids cannot move beyond the MTOC. Whether this is because they fail to switch transport direction or were unable to dock at the nucleus remains open. Alternatively, capsids could bind nuclear transport factors at the NPC. In that case, ΔNLS capsids would migrate from the MTOC to the NPC but, due to failure to engage the NPC, inadvertently accumulate near the MTOC. The data from my experiments imply that the inserted NLS sequences were sufficient to allow docking but cannot distinguish between these two scenarios.

7.2.3 Wider impact of NLS study

Although resistance to conventional anti‐virals including acyclovir is rare, in immunocompromised patients the risk of generating resistant virus is increased. Thus, it is important to develop new drugs targeting conserved cellular features of virus-host interactions, which may be exploited to lower the risk of emerging resistance. One possible target is the interaction of capsids with their cognate nuclear transport factor. The presented results and tools could have a wider impact on the HV community. Using the knowledge about conservation of the NLS in all HVs and its ability to rescue HSV-1 replication the chimeric VP1-2 derivatives could be used to elucidate the nuclear entry mechanism of potentially all HVs and guide the rational design of inhibitors of capsid nuclear docking. Using a combination of nuclear localisation of chimeric NT6 (or β-gal fusions), recombinant HSV-1 capsid binding studies and HSV-1 uncoating assays, it is possible to directly measure the contributions of individual proteins to the docking process by siRNA-mediated knockdown, chemical inhibition or expression of dominant negative nuclear transport factors. Using the developed HSV-1 based system has a few notable advantages over studying the native virus.

- 196 -

General discussion

(i) Very little cloning and prior knowledge about structure of the VP1-2 homologues in questions are required (ii) Nuclear import assays can easily be expanded with further NLSs and the constructs could form the basis for interaction studies (iii) HSV-1 replicates to reasonably high titre in many cell lines (e.g. RSC, BHK) even in the presence of a foreign NLS (iv) The one-step capsid isolation protocol can be performed on any high titre HSV-1 stock (v) HSV-1 DNA can be labelled with ‘clickable’ nucleoside analogues such as EdC (Dembowski & DeLuca 2015) (vi) Our group has developed a HSV-1 genome uncoating assay which can be performed in susceptible cell lines. For this assay HSV-1 genomes are labelled with ‘clickable’ nucleoside analogues. Labelled genomes can in turn be detected in newly infected nuclei.

In future work using techniques established in this work and in combination with other tools, a specific workflow could be established. Initially, nuclear localisation of NT6 could be scored in the presence of specific inhibitors, siRNAs or dominant negative importin-αs. Secondly, the effect of inhibition of certain factors could subsequently be verified in a genome uncoating assay using the chimeric viruses. Finally, using the one-step capsid purification nuclear transport factor interactions could be confirmed. This could lead to the identification of nuclear transport factors used by each HV and be used for the rational design of drugs to inhibit HV infection.

7.2.4 Some intriguing questions remain 7.2.4.1 NLS function

Nuclear localisation is a complicated process which appears to not only involve nuclear transport factors. There are several mechanisms for a cell to modulate the function of a protein which act at the level of protein expression or stability. Additionally, protein function can also be modulated at the level of localisation, more specifically by inhibiting or augmenting NLS activity through, for example, phosphorylation (Jain et al. 1993; Flanagan et al. 1991). Since NLS strength had no influence on ΔNLS rescue efficiency for the homologous NLSs, it would be intriguing to test if the functional differences could be precipitated by altered regulatability. The lack of ΔNLS rescue using the HHV-8 motif and the reduction in viral fitness seen with the motifs derived from the α-, β- and γ1-HVs also led to another hypothesis. The sub-group specific organisation and conservation indicate that this motif might be adapted to the viral life cycle,

- 197 -

General discussion specifically the cell types these viruses infect. This could be a reason why in Vero cells all the homologous NLSs were essentially non-functional. The Vero cell defect could prove a fortuitous finding, since it enables the investigation of the observed phenotype by qualitative (microscopy) and quantitative (protein expression time courses, qRT-PCR) means. Effectively, the contribution of cellular (nuclear transport) factors to the HV capsid entry process can be assessed when individual nuclear transport factors are exogenously overexpressed. On a grander scale the linker regions between USP and the body of VP1-2 are highly heterogenous even within sub-families with only the NLS region being reasonably well conserved. More specifically, only the basic residues of R4 were highly conserved throughout the sub-families (Figure 8). In contrast, R5 (second basic patch) or the spacer exhibits very little resemblance between HV sub-families except between very closely related species (i.e. HSV-1/2 or EBV/McHV-4). It appears as though a comparatively high selective pressure acted upon the first basic patch compared to the rest of the linker region (i.e. R5 or the residues surrounding R1) suggesting that they might bind similarly conserved cellular proteins in their respective hosts (e.g. major binding site of α-importins). In contrast, the surrounding area could fulfil more plastic accessory functions (e.g. presentation or regulation of NLS function in cell/host-specific manner, proteolytic site, etc.). This conclusion may not be that far fetched since the basic residues in R4 contribute most of the function of this region (see K428R and K428A defects). Future work to refine the system to co-precipitate VP1-2 interacting proteins could lead to elucidation of the role of the more variable parts of R1.

7.2.4.2 The expanding roles of VP1-2

In the third results chapter I attempted confirm interaction of VP1-2 with DTX3L. There were several more proteins identified by MS that were not followed up but their functions open up new opportunities to expand the list of roles for VP1-2 in the virus life cycle that could explain some of the previously unexplained defects observed with USP mutants of PRV or the temperature sensitive HSV-1 mutant tsB7 (Böttcher et al. 2008; Abaitua et al. 2011; Abaitua et al. 2009). It may also not be surprising that NT3 gave rise to such an extensive list of interacting proteins since the region following the USP (including the NLS) likely assumes no rigid secondary structure. This plasticity could potentially enable VP1-2 to interact with many different proteins as was suggested for intrinsically disordered regions (Dunker et al. 2001 and references therein); these binding proteins could in turn be modulated by the USP. Since this study revealed a previously unappreciated (potential) function of VP1-2 the question remains how modulating DTX3L (and/or PARP9) activity could be beneficial to virus replication. This intriguing protein has previously been shown to affect endosomal sorting (Holleman & Marchese 2014)

- 198 -

General discussion and the DDR (Yan et al. 2009; Yan et al. 2013), processes which are extensively modulated by HSV-1 (some aspects discussed in 5.3.1.4). Hypothetically, VP1-2 can act as a soluble, potentially truncated (Kattenhorn et al. 2005) or capsid-bound full length protein (see sections 1.2.2 and 1.2.3) during viral replication. But in which form does VP1-2 affect DTX3L? Two equally exciting possibilities arise. (i) The soluble full length or N-terminal VP1-2 product (i.e. the USP) could affect the DDR late on during infection to facilitate DNA replication. The physiological VP1-2 cleavage product described by Kattenhorn et al. (2005) differs from NT3 in that NT3 is approximately 400 residues longer and thus could contain regions for interaction with DTX3L downstream of the cleavage site. Additionally, it is unknown whether the physiological cleavage product of VP1-2 comprises the NLS and thus whether NT3, which mostly localises to the nucleus, is a relevant system. Follow up experiments (discussed earlier) are underway to establish the nature of the VP1-2-DTX3L interaction and if this augments viral replication through modulation of the DDR. (ii) VP1-2, in its full length capsid-bound form, facilitates secondary envelopment of HV capsids and is required for capsid movement (see section 1.2.3). The ESCRT pathway of endosomal sorting is extensively regulated by ubiquitination (Urbé 2005). The modulation of the function of the E3 ubiquitin ligase, DTX3L, by the VP1-2 USP in one pathway (DDR) could be reason to believe that the endosomal sorting functions of DTX3L (Holleman & Marchese 2014) are equally affected. These data could link functions of VP1-2 during different life cycle stages to the USP and form the basis of new studies to unravelling open questions (e.g. the mechanism of the defect in egress of the PRV lacking VP1-2 USP function found by Böttcher et al. 2008).

- 199 -

References

8 References

Abaitua, F., Daikoku, T., Crump, C.M., Bolstad, M. & O’Hare, P., 2011. A single mutation responsible for temperature-sensitive entry and assembly defects in the VP1-2 protein of herpes simplex virus. Journal of virology, 85(5), pp.2024–2036. Abaitua, F., Hollinshead, M., Bolstad, M., Crump, C.M. & O’Hare, P., 2012. A Nuclear Localization Signal in Herpesvirus Protein VP1-2 Is Essential for Infection via Capsid Routing to the Nuclear Pore. Journal of Virology, 86(17), pp.8998–9014. Abaitua, F. & O’Hare, P., 2008. Identification of a highly conserved, functional nuclear localization signal within the N-terminal region of herpes simplex virus type 1 VP1-2 tegument protein. Journal of virology, 82(11), pp.5234–5244. Abaitua, F., Souto, R.N., Browne, H., Daikoku, T. & O’Hare, P., 2009. Characterization of the herpes simplex virus (HSV)-1 tegument protein VP1-2 during infection with the HSV temperature-sensitive mutant tsB7. Journal of General Virology, 90(10), pp.2353–2363. Abaitua, F., Zia, F.R., Hollinshead, M. & O’Hare, P., 2013. Polarized cell migration during cell-to-cell transmission of herpes simplex virus in human skin keratinocytes. Journal of virology, 87(14), pp.7921–32. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3700176&tool=pmcentrez&rendertype=abstract. Ace, C.I., Dalrymple, M.A., Ramsay, F.H., Preston, V.G. & Preston, C.M., 1988. Mutational analysis of the herpes simplex virus type 1 transinducing factor Vmw65. The Journal of general virology, 69 ( Pt 10, pp.2595–2605. Adamson, W.E., Mcnab, D., Preston, V.G. & Rixon, F.J., 2006. Mutational Analysis of the Herpes Simplex Virus Triplex Protein VP19C Mutational Analysis of the Herpes Simplex Virus Triplex Protein VP19C. Journal of Virology, 80(3), pp.1537–1548. Aggarwal, A., Miranda-Saksena, M., Boadle, R.A., Kelly, B.J., Diefenbach, R.J., Alam, W. & Cunningham, A.L., 2012. Ultrastructural Visualization of Individual Tegument Protein Dissociation during Entry of Herpes Simplex Virus 1 into Human and Rat Dorsal Root Ganglion Neurons. Journal of Virology, 86(11), pp.6123–6137. Aglipay, J.A., Martin, S.A., Tawara, H., Lee, S.W. & Ouchi, T., 2006. ATM activation by ionizing radiation requires BRCA1-associated BAAT1. Journal of Biological Chemistry, 281(14), pp.9710–9718. Aitchison, J.D., Blobel, G. & Rout, M.P., 1996. Kap104p: a karyopherin involved in the nuclear transport of messenger RNA binding proteins. Science (New York, N.Y.), 274(5287), pp.624–627. Akey, C.W. & Goldfarb, D.S., 1989. Protein import through the nuclear pore complex is a multistep process. Journal of Cell Biology, 109(3), pp.971–982. Alonso-Caplen, F. V, Nemeroff, M.E., Qiu, Y. & Krug, R.M., 1992. Nucleocytoplasmic transport: the influenza virus NS1 protein regulates the transport of spliced NS2 mRNA and its precursor NS1 mRNA. Genes & development, 6(2), pp.255–67. Available at: http://www.ncbi.nlm.nih.gov/pubmed/1531330 [Accessed April 12, 2015]. Anderson, F., Savulescu, A.F., Rudolph, K., Schipke, J., Cohen, I., Ibiricu, I., Rotem, A., Grünewald, K., Sodeik, B. & Harel, A., 2014. Targeting of Viral Capsids to Nuclear Pores in a Cell-Free Reconstitution System. Traffic, 15, pp.1266–1281. Available at: http://doi.wiley.com/10.1111/tra.12209. Anderson, H. a, Chen, Y. & Norkin, L.C., 1996. Bound simian virus 40 translocates to caveolin-enriched membrane domains, and its entry is inhibited by drugs that selectively disrupt caveolae. Molecular biology of the cell, 7(11), pp.1825–1834. Antinone, S.E., Shubeita, G.T., Coller, K.E., Lee, J.I., Haverlock-Moyns, S., Gross, S.P. & Smith, G. a, 2006. The Herpesvirus capsid surface protein, VP26, and the majority of the tegument proteins are dispensable for capsid transport toward the nucleus. Journal of virology, 80(11), pp.5494–5498. Ao, Z., Danappa Jayappa, K., Wang, B., Zheng, Y., Kung, S., Rassart, E., Depping, R., Kohler, M., Cohen, E. a & Yao, X., 2010. Importin alpha3 interacts with HIV-1 integrase and contributes to HIV-1 nuclear import and replication. Journal of virology, 84(17), pp.8650–8663. Ao, Z., Huang, G., Yao, H., Xu, Z., Labine, M., Cochrane, A.W. & Yao, X., 2007. Interaction of human immunodeficiency virus type 1 integrase with cellular nuclear import receptor importin 7 and its impact on viral replication. Journal of Biological Chemistry, 282(18), pp.13456– 13467. Ao, Z., Jayappa, K.D., Wang, B., Zheng, Y., Wang, X., Peng, J. & Yao, X., 2012. Contribution of host nucleoporin 62 in HIV-1 integrase chromatin association and viral DNA integration. Journal of Biological Chemistry, 287(13), pp.10544–10555. Atasheva, S., Fish, A., Fornerod, M. & Frolova, E.I., 2010. Venezuelan equine Encephalitis virus capsid protein forms a tetrameric complex with CRM1 and importin alpha/beta that obstructs nuclear pore complex function. Journal of virology, 84(9), pp.4158–4171. Atasheva, S., Garmashova, N., Frolov, I. & Frolova, E., 2008. Venezuelan equine encephalitis virus capsid protein inhibits nuclear import in Mammalian but not in mosquito cells. Journal of virology, 82(8), pp.4028–4041. Au, S. & Panté, N., 2012. Nuclear transport of baculovirus: revealing the nuclear pore complex passage. Journal of structural biology, 177(1), pp.90–8. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22100338 [Accessed October 30, 2014]. Au, S., Wu, W. & Panté, N., 2013. Baculovirus nuclear import: Open, nuclear pore complex (NPC) sesame. Viruses, 5(7), pp.1885–1900. Avitabile, E., Di Gaeta, S., Torrisi, M.R., Ward, P.L., Roizman, B. & Campadelli-Fiume, G., 1995. Redistribution of microtubules and Golgi apparatus in herpes simplex virus-infected cells and their role in viral exocytosis. Journal of virology, 69(12), pp.7472–7482. Babcock, H.P., Chen, C. & Zhuang, X., 2004. Using single-particle tracking to study nuclear trafficking of viral genes. Biophysical journal, 87(4), pp.2749–2758. Available at: http://dx.doi.org/10.1529/biophysj.104.042234. Baker, M.L., Jiang, W., Rixon, F.J. & Chiu, W., 2005. Common ancestry of herpesviruses and tailed DNA bacteriophages. Journal of virology, 79(23), pp.14967–14970. Bardina, M. V, Lidsky, P. V, Sheval, E. V, Fominykh, K. V, van Kuppeveld, F.J.M., Polyakov, V.Y. & Agol, V.I., 2009. Mengovirus-induced

- 200 -

References

rearrangement of the nuclear pore complex: hijacking cellular phosphorylation machinery. Journal of virology, 83(7), pp.3150–3161. Batterson, W., Furlong, D. & Roizman, B., 1983. Molecular Genetics of Herpes Simplex Virus VIII . Further Characterization of a Temperature- Sensitive Mutant Reproductive Cycle. Journal of virology, 45(1), pp.397–407. Batterson, W. & Roizman, B., 1983. Characterization of the herpes simplex virion-associated factor responsible for the induction of alpha genes. Journal of virology, 46(2), pp.371–377. Bayliss, R., Leung, S.W., Baker, R.P., Quimby, B.B., Corbett, A.H. & Stewart, M., 2002. Structural basis for the interaction between NTF2 and nucleoporin FxFG repeats. EMBO Journal, 21(12), pp.2843–2853. Bednenko, J., Cingolani, G. & Gerace, L., 2003. Nucleocytoplasmic Transport: Navigating the Channel. Traffic, 4(3), pp.127–135. Available at: http://doi.wiley.com/10.1034/j.1600-0854.2003.00109.x. Belov, G.A., Lidsky, P. V, Mikitas, O. V, Egger, D., Lukyanov, K.A., Bienz, K. & Agol, V.I., 2004. Bidirectional increase in permeability of nuclear envelope upon poliovirus infection and accompanying alterations of nuclear pores. Journal of virology, 78(18), pp.10166–10177. Bhoj, V.G. & Chen, Z.J., 2009. Ubiquitylation in innate and adaptive immunity. Nature, 458(March), pp.430–437. Bischoff, F.R. & Görlich, D., 1997. RanBP1 is crucial for the release of RanGTP from importin β-related nuclear transport factors. FEBS Letters, 419(2-3), pp.249–254. Bjerke, S.L. & Roller, R.J., 2006. Roles for herpes simplex virus type 1 UL34 and US3 proteins in disrupting the nuclear lamina during herpes simplex virus type 1 egress. Virology, 347(2), pp.261–76. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2993110&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Blom, N., Gammeltoft, S. & Brunak, S., 1999. Sequence and structure-based prediction of eukaryotic protein phosphorylation sites. Journal of molecular biology, 294(5), pp.1351–1362. Blouin, a & Blaho, J. a, 2001. Assessment of the subcellular localization of the herpes simplex virus structural protein VP22 in the absence of other viral gene products. Virus research, 81(1-2), pp.57–68. Bohannon, K.P., Jun, Y., Gross, S.P. & Smith, G.A., 2013. Differential protein partitioning within the herpesvirus tegument and envelope underlies a complex and variable virion architecture. Proceedings of the National Academy of Sciences of the United States of America, 110, pp.E1613–20. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3637715&tool=pmcentrez&rendertype=abstract. La Boissière, S., Hughes, T. & O'Hare, P., 1999. HCF-dependent nuclear import of VP16. EMBO Journal, 18(2), pp.480–489. Boissière, S. La, Izeta, A., Malcomber, S. & O’Hare, P.F., 2004. Compartmentalization of VP16 in Cells Infected with Recombinant Herpes Simplex Virus Expressing VP16-Green Fluorescent Protein Fusion Proteins. Journal of virology, 78(15), pp.8002–8014. Bolstad, M., Abaitua, F., Crump, C.M. & O’Hare, P., 2011. Autocatalytic Activity of the Ubiquitin-Specific Protease Domain of Herpes Simplex Virus 1 VP1-2. Journal of virology, 85(17), pp.8738–8751. Böttcher, S., Klupp, B.G., Granzow, H., Fuchs, W., Michael, K. & Mettenleiter, T.C., 2006. Identification of a 709-amino-acid internal nonessential region within the essential conserved tegument protein (p)UL36 of pseudorabies virus. Journal of virology, 80(19), pp.9910–9915. Böttcher, S., Maresch, C., Granzow, H., Klupp, B.G., Teifke, J.P. & Mettenleiter, T.C., 2008. Mutagenesis of the active-site cysteine in the ubiquitin-specific protease contained in large tegument protein pUL36 of pseudorabies virus impairs viral replication in vitro and neuroinvasion in vivo. Journal of virology, 82(12), pp.6009–6016. Brachner, A. & Foisner, R., 2011. Evolvement of LEM proteins as chromatin tethers at the nuclear periphery. Biochemical Society Transactions, 39(6), pp.1735–1741. Brameier, M., Krings, A. & MacCallum, R.M., 2007. NucPred--predicting nuclear localization of proteins. Bioinformatics (Oxford, England), 23(9), pp.1159–1160. Brock, I., Krüger, M., Mertens, T. & von Einem, J., 2013. Nuclear targeting of human cytomegalovirus large tegument protein pUL48 is essential for viral growth. Journal of virology, 87(10), pp.6005–19. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3648141&tool=pmcentrez&rendertype=abstract. Brown, J.C. & Newcomb, W.W., 2011. Herpesvirus capsid assembly: insights from structural analysis. Current opinion in virology, 1(2), pp.142– 9. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3171831&tool=pmcentrez&rendertype=abstract [Accessed April 17, 2015]. Browne, H., Bruun, B. & Minson, T., 2001. Plasma membrane requirements for cell fusion induced by herpes simplex virus type 1 glycoproteins gB, gD, gH and gL. J. Gen. Virol., 82(6), pp.1419–1422. Available at: http://vir.sgmjournals.org/content/82/6/1419.full [Accessed March 14, 2015]. Bucks, M. a, Murphy, M. a, O’Regan, K.J. & Courtney, R.J., 2011. Identification of interaction domains within the UL37 tegument protein of herpes simplex virus type 1. Virology, 416(1-2), pp.42–53. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3617580&tool=pmcentrez&rendertype=abstract [Accessed October 21, 2014]. Bucks, M. a., O’Regan, K.J., Murphy, M. a., Wills, J.W. & Courtney, R.J., 2007. Herpes simplex virus type 1 tegument proteins VP1/2 and UL37 are associated with intranuclear capsids. Virology, 361(2), pp.316–324. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2710585&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Bui, M., Whittaker, G. & Helenius, a, 1996. Effect of M1 protein and low pH on nuclear transport of influenza virus ribonucleoproteins.

- 201 -

References

Journal of virology, 70(12), pp.8391–8401. Bukrinsky, M.I., Haggerty, S., Dempsey, M.P., Sharova, N., Adzhubel, A., Spitz, L., Lewis, P., Goldfarb, D., Emerman, M. & Stevenson, M., 1993. A nuclear localization signal within HIV-1 matrix protein that governs infection of non-dividing cells. Nature, 365(6447), pp.666–669. Bullido, R., Gómez-Puertas, P., Albo, C. & Portela, A., 2000. Several protein regions contribute to determine the nuclear and cytoplasmic localization of the influenza A virus nucleoprotein. Journal of General Virology, 81(1), pp.135–142. Bullough, P.A., Hughson, F.M., Skehel, J.J. & Wiley, D.C., 1994. Structure of influenza haemagglutinin at the pH of membrane fusion. Nature, 371(6492), pp.37–43. Bunting, S.F., Callén, E., Wong, N., Chen, H.T., Polato, F., Gunn, A., Bothmer, A., Feldhahn, N., Fernandez-Capetillo, O., Cao, L., Xu, X., Deng, C.X., Finkel, T., Nussenzweig, M., Stark, J.M. & Nussenzweig, A., 2010. 53BP1 inhibits homologous recombination in brca1-deficient cells by blocking resection of DNA breaks. Cell, 141(2), pp.243–254. Burch, A.D. & Weller, S.K., 2004. Nuclear Sequestration of Cellular Chaperone and Proteasomal Machinery during Herpes Simplex Virus Type 1 Infection Nuclear Sequestration of Cellular Chaperone and Proteasomal Machinery during Herpes Simplex Virus Type 1 Infection. Journal of virology, 78(13), p.7175. Burgdorf, S., Leister, P. & Scheidtmann, K.H., 2004. TSG101 Interacts with Apoptosis-antagonizing Transcription Factor and Enhances Androgen Receptor-mediated Transcription by Promoting Its Monoubiquitination. Journal of Biological Chemistry, 279(17), pp.17524– 17534. Bürkle, A. & Virág, L., 2013. Poly(ADP-ribose): PARadigms and PARadoxes. Molecular Aspects of Medicine, 34(6), pp.1046–1065. Buser, C., Walther, P., Mertens, T. & Michel, D., 2007. Cytomegalovirus primary envelopment occurs at large infoldings of the inner nuclear membrane. Journal of virology, 81(6), pp.3042–8. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1865996&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Butin-Israeli, V., Ben-nun-Shaul, O., Kopatz, I., Adam, S.A., Shimi, T., Goldman, R.D. & Oppenheim, A., 2011. Simian virus 40 induces lamin A/C fluctuations and nuclear envelope deformation during cell entry. Nucleus, 2(4), pp.320–330. Butin-Israeli, V., Drayman, N. & Oppenheim, A., 2010. Simian virus 40 infection triggers a balanced network that includes apoptotic, survival, and stress pathways. Journal of virology, 84(7), pp.3431–3442. Calistri, a, Munegato, D., Toffoletto, M., Celestino, M., Franchin, E., Comin, A., Sartori, E., Salata, C., Parolin, C. & Palù, G., 2014. Functional interaction between the ESCRT-I component TSG101 and the HSV-1 tegument ubiquitin specific protease. Journal of Cellular Physiology, (June), p.n/a–n/a. Available at: http://doi.wiley.com/10.1002/jcp.24890. Campadelli-Fiume, G., Cocchi, F., Menotti, L. & Lopez, M., 2000. The novel receptors that mediate the entry of herpes simplex viruses and animal alphaherpesviruses into cells. Reviews in Medical Virology, 10(5), pp.305–319. Campadelli-fiume, G., Farabegoli, F., Gaeta, S.D.I. & Roizman, B., 1991. Origin of Unenveloped Capsids in the Cytoplasm of Cells Infected with Herpes Simplex Virus 1. Journal of virology, 65(3), pp.1589–1595. Campbell, M.E., Palfreyman, J.W. & Preston, C.M., 1984. Identification of herpes simplex virus DNA sequences which encode a trans-acting polypeptide responsible for stimulation of immediate early transcription. Journal of molecular biology, 180(1), pp.1–19. Cano-Monreal, G.L., Wylie, K.M., Cao, F., Tavis, J.E. & Morrison, L. a., 2009. Herpes simplex virus 2 UL13 protein kinase disrupts nuclear lamins. Virology, 392(1), pp.137–147. Available at: http://dx.doi.org/10.1016/j.virol.2009.06.051. Capelson, M., Liang, Y., Schulte, R., Mair, W., Wagner, U. & Hetzer, M.W., 2010. Chromatin-Bound Nuclear Pore Components Regulate Gene Expression in Higher Eukaryotes. Cell, 140(3), pp.372–383. Available at: http://dx.doi.org/10.1016/j.cell.2009.12.054. Cardone, G., Newcomb, W.W., Cheng, N., Wingfield, P.T., Trus, B.L., Brown, J.C. & Steven, A.C., 2012. The UL36 Tegument Protein of Herpes Simplex Virus 1 Has a Composite Binding Site at the Capsid Vertices. Journal of Virology, 86(8), pp.4058–4064. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3318633&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Casjens, S. & King, J., 1975. Virus assembly. Annual review of biochemistry, 44, pp.555–611. Chan, E.Y., Qian, W.-J., Diamond, D.L., Liu, T., Gritsenko, M.A., Monroe, M.E., Camp, D.G., Smith, R.D. & Katze, M.G., 2007. Quantitative analysis of human immunodeficiency virus type 1-infected CD4+ cell proteome: dysregulated cell cycle progression and nuclear transport coincide with robust virus production. Journal of virology, 81(14), pp.7571–7583. Chan, E.Y., Sutton, J.N., Jacobs, J.M., Bondarenko, A., Smith, R.D. & Katze, M.G., 2009. Dynamic host energetics and cytoskeletal proteomes in human immunodeficiency virus type 1-infected human primary CD4 cells: analysis by multiplexed label-free mass spectrometry. Journal of virology, 83(18), pp.9283–9295. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2738255&tool=pmcentrez&rendertype=abstract [Accessed April 12, 2015]. Chang, C.-W., Couñago, R.M., Williams, S.J., Boden, M. & Kobe, B., 2014. The distribution of different classes of nuclear localization signals (NLSs) in diverse organisms and the utilization of the minor NLS-binding site inplantnuclear import factor importin-α. Plant Signaling & Behavior, 8(10), p.e25976. Available at: http://www.tandfonline.com/doi/abs/10.4161/psb.25976. Chang, C.-W., Lee, C.-P., Su, M.-T., Tsai, C.-H. & Chen, M.-R., 2015. BGLF4 kinase modulates the structure and transport preference of the nuclear pore complex to facilitate nuclear import of Epstein-Barr virus lytic proteins. Journal of virology, 89(3), pp.1703–18. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4300756&tool=pmcentrez&rendertype=abstract [Accessed April 12, 2015]. Chen, C. & Okayama, H., 1987. High-efficiency transformation of mammalian cells by plasmid DNA. Molecular and cellular biology, 7(8),

- 202 -

References

pp.2745–2752. Chen, C. & Zhuang, X., 2008. Epsin 1 is a cargo-specific adaptor for the clathrin-mediated endocytosis of the influenza virus. Proceedings of the National Academy of Sciences of the United States of America, 105(33), pp.11790–11795. Chen, M.S., Summers, W.P., Walker, J., Summers, W.C. & Prusoff, W.H., 1979. Characterization of pyrimidine deoxyribonucleoside kinase (thymidine kinase) and thymidylate kinase as a multifunctional enzyme in cells transformed by herpes simplex virus type 1 and in cells infected with mutant strains of herpes simplex virus. Journal of virology, 30(3), pp.942–945. Chen, Z., Li, Y. & Krug, R.M., 1999. Influenza A virus NS1 protein targets poly(A)-binding protein II of the cellular 3’-end processing machinery. The EMBO journal, 18(8), pp.2273–83. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1171310&tool=pmcentrez&rendertype=abstract [Accessed March 24, 2015]. Cheng, G., Brett, M.-E. & He, B., 2002. Signals that dictate nuclear, nucleolar, and cytoplasmic shuttling of the gamma(1)34.5 protein of herpes simplex virus type 1. Journal of virology, 76(18), pp.9434–9445. Chou, K.C. & Shen, H. Bin, 2010. A new method for predicting the subcellular localization of eukaryotic proteins with both single and multiple sites: Euk-mPLoc 2.0. PLoS ONE, 5(4). Chou, P.Y. & Fasman, G.D., 1974. Conformational parameters for amino acids in helical, β-sheet, and random coil regions calculated from proteins. Biochemistry, 13(2), pp.211–222. Chou, Y. ying, Heaton, N.S., Gao, Q., Palese, P., Singer, R. & Lionnet, T., 2013. Colocalization of Different Influenza Viral RNA Segments in the Cytoplasm before Viral Budding as Shown by Single-molecule Sensitivity FISH Analysis. PLoS Pathogens, 9(5). Christ, F., Thys, W., De Rijck, J., Gijsbers, R., Albanese, A., Arosio, D., Emiliani, S., Rain, J.C., Benarous, R., Cereseto, A. & Debyser, Z., 2008. Transportin-SR2 Imports HIV into the Nucleus. Current Biology, 18(16), pp.1192–1202. Chung, K.M., Lee, J., Kim, J.E., Song, O.K., Cho, S., Lim, J., Seedorf, M., Hahm, B. & Jang, S.K., 2000. Nonstructural protein 5A of hepatitis C virus inhibits the function of karyopherin beta3. Journal of virology, 74(11), pp.5233–5241. Cingolani, G., Petosa, C., Weis, K. & Müller, C.W., 1999. Structure of importin-beta bound to the IBB domain of importin-alpha. Nature, 399(6733), pp.221–229. Clever, J., Yamada, M. & Kasamatsu, H., 1991. Import of simian virus 40 virions through nuclear pore complexes. Proceedings of the National Academy of Sciences of the United States of America, 88(16), pp.7333–7337. Cockrell, S.K., Huffman, J.B., Toropova, K., Conway, J.F. & Homa, F.L., 2011. Residues of the UL25 protein of herpes simplex virus that are required for its stable interaction with capsids. Journal of virology, 85(10), pp.4875–4887. Cohen, S., Behzad, A.R., Carroll, J.B. & Panté, N., 2006. Parvoviral nuclear import: Bypassing the host nuclear-transport machinery. Journal of General Virology, 87(11), pp.3209–3213. Cohen, S., Marr, A.K., Garcin, P. & Panté, N., 2011. Nuclear envelope disruption involving host caspases plays a role in the parvovirus replication cycle. Journal of virology, 85(10), pp.4863–4874. Cohen, S. & Panté, N., 2005. Pushing the envelope: Microinjection of Minute virus of mice into Xenopus oocytes causes damage to the nuclear envelope. Journal of General Virology, 86(12), pp.3243–3252. Coller, K.E., Lee, J.I.-H., Ueda, A. & Smith, G.A., 2007. The capsid and tegument of the alphaherpesviruses are linked by an interaction between the UL25 and VP1/2 proteins. Journal of virology, 81(21), pp.11790–11797. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2168758&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Conti, E. & Kuriyan, J., 2000. Crystallographic analysis of the specific yet versatile recognition of distinct nuclear localization signals by karyopherin ?? Structure, 8(3), pp.329–338. Conti, E., Müller, C.W. & Stewart, M., 2006. Karyopherin flexibility in nucleocytoplasmic transport. Current Opinion in Structural Biology, 16(2), pp.237–244. Copeland, A.M., Newcomb, W.W. & Brown, J.C., 2009. Herpes simplex virus replication: roles of viral proteins and nucleoporins in capsid- nucleus attachment. Journal of virology, 83(4), pp.1660–1668. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2643781&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. La Cour, T., Kiemer, L., Mølgaard, A., Gupta, R., Skriver, K. & Brunak, S., 2004. Analysis and prediction of leucine-rich nuclear export signals. Protein Engineering, Design and Selection, 17(6), pp.527–536. Cronshaw, J.M., Krutchinsky, A.N., Zhang, W., Chait, B.T. & Matunis, M.J., 2002. Proteomic analysis of the mammalian nuclear pore complex. The Journal of cell biology, 158(5), pp.915–927. Cros, J.F., García-Sastre, A. & Palese, P., 2005. An unconventional NLS is critical for the nuclear import of the influenza A virus nucleoprotein and ribonucleoprotein. Traffic, 6(3), pp.205–213. Crump, C.M., Yates, C. & Minson, T., 2007. Herpes simplex virus type 1 cytoplasmic envelopment requires functional Vps4. Journal of virology, 81(14), pp.7380–7387. Cutress, M.L., Whitaker, H.C., Mills, I.G., Stewart, M. & Neal, D.E., 2008. Structural basis for the nuclear import of the human androgen receptor. Journal of cell science, 121(Pt 7), pp.957–968. Dargan, D.J., Patel, a H. & Subak-Sharpe, J.H., 1995. PREPs: herpes simplex virus type 1-specific particles produced by infected cells when viral DNA replication is blocked. Journal of virology, 69(8), pp.4924–4932. Darlington, R.W. & Moss, L.H., 1968. Herpesvirus envelopment. Journal of virology, 2(1), pp.48–55.

- 203 -

References

Davison, A.J., 2002. Evolution of the herpesviruses. Veterinary Microbiology, 86(1-2), pp.69–88. Dechat, T., Pfleghaar, K., Sengupta, K., Shimi, T., Shumaker, D.K., Solimando, L. & Goldman, R.D., 2008. Nuclear lamins: Major factors in the structural organization and function of the nucleus and chromatin. Genes and Development, 22(7), pp.832–853. Dembowski, J. a. & DeLuca, N. a., 2015. Selective Recruitment of Nuclear Factors to Productively Replicating Herpes Simplex Virus Genomes. PLOS Pathogens, 11(5), p.e1004939. Available at: http://dx.plos.org/10.1371/journal.ppat.1004939. Deng, T., Engelhardt, O.G., Thomas, B., Akoulitchev, A. V, Brownlee, G.G. & Fodor, E., 2006. Role of ran binding protein 5 in nuclear import and assembly of the influenza virus RNA polymerase complex. Journal of virology, 80(24), pp.11911–11919. Denning, D.P., Patel, S.S., Uversky, V., Fink, A.L. & Rexach, M., 2003. Disorder in the nuclear pore complex: the FG repeat regions of nucleoporins are natively unfolded. Proceedings of the National Academy of Sciences of the United States of America, 100(5), pp.2450–2455. Desai, P., DeLuca, N. a & Person, S., 1998. Herpes simplex virus type 1 VP26 is not essential for replication in cell culture but influences production of infectious virus in the nervous system of infected mice. Virology, 247, pp.115–124. Desai, P., Sexton, G.L., McCaffery, J.M. & Person, S., 2001. A null mutation in the gene encoding the herpes simplex virus type 1 UL37 polypeptide abrogates virus maturation. Journal of virology, 75(21), pp.10259–10271. Desai, P.J., 2000. A null mutation in the UL36 gene of herpes simplex virus type 1 results in accumulation of unenveloped DNA-filled capsids in the cytoplasm of infected cells. Journal of virology, 74(24), pp.11608–11618. Diacumakos, E.G. & Gershey, E.L., 1977. Uncoating and gene expression of simian virus 40 in CV-1 cell nuclei inoculated by microinjection. Journal of virology, 24(3), pp.903–906. Dias, S.M.G., Wilson, K.F., Rojas, K.S., Ambrosio, A.L.B. & Cerione, R. a, 2009. The molecular basis for the regulation of the cap-binding complex by the importins. Nature structural & molecular biology, 16(9), pp.930–937. Available at: http://dx.doi.org/10.1038/nsmb.1649. Dingwall, C. & Laskey, R.A., 1991. Nuclear targeting sequences — a consensus? Trends in Biochemical Sciences, 16, pp.478–481. Dingwall, C., Sharnick, S. V & Laskey, R.A., 1982. A polypeptide domain that specifies migration of nucleoplasmin into the nucleus. Cell, 30(2), pp.449–458. Döhner, K., Nagel, C.H. & Sodeik, B., 2005. Viral stop-and-go along microtubules: Taking a ride with dynein and kinesins. Trends in Microbiology, 13(7), pp.320–327. Döhner, K., Radtke, K., Schmidt, S. & Sodeik, B., 2006. Eclipse phase of herpes simplex virus type 1 infection: Efficient dynein-mediated capsid transport without the small capsid protein VP26. Journal of virology, 80(16), pp.8211–8224. Donnelly, M. & Elliott, G., 2001. Nuclear localization and shuttling of herpes simplex virus tegument protein VP13/14. Journal of virology, 75(6), pp.2566–2574. Douglas, M.W., Diefenbach, R.J., Homa, F.L., Miranda-Saksena, M., Rixon, F.J., Vittone, V., Byth, K. & Cunningham, A.L., 2004. Herpes simplex virus type 1 capsid protein VP26 interacts with dynein light chains RP3 and Tctex1 and plays a role in retrograde cellular transport. Journal of Biological Chemistry, 279(27), pp.28522–28530. Dunker, A.K., Lawson, J.D., Brown, C.J., Williams, R.M., Romero, P., Oh, J.S., Oldfield, C.J., Campen, A.M., Ratliff, C.M., Hipps, K.W., Ausio, J., Nissen, M.S., Reeves, R., Kang, C., Kissinger, C.R., Bailey, R.W., Griswold, M.D., Chiu, W., Garner, E.C. & Obradovic, Z., 2001. Intrinsically disordered protein. Journal of Molecular Graphics and Modelling, 19(1), pp.26–59. Dworetzky, S.I., Lanford, R.E. & Feldherr, C.M., 1988. The effects of variations in the number and sequence of targeting signals on nuclear uptake. Journal of Cell Biology, 107(4), pp.1279–1287. Ebina, H., Aoki, J., Hatta, S., Yoshida, T. & Koyanagi, Y., 2004. Role of Nup98 in nuclear entry of human immunodeficiency virus type 1 cDNA. Microbes and Infection, 6(8), pp.715–724. Eierhoff, T., Hrincius, E.R., Rescher, U., Ludwig, S. & Ehrhardt, C., 2010. The epidermal growth factor receptor (EGFR) promotes uptake of influenza a viruses (IAV) into host cells. PLoS Pathogens, 6(9). von Einem, J., Schumacher, D., O’Callaghan, D.J. & Osterrieder, N., 2006. The alpha-TIF (VP16) homologue (ETIF) of equine herpesvirus 1 is essential for secondary envelopment and virus egress. Journal of virology, 80(6), pp.2609–2620. Elliott, G. & O’Hare, P., 2000. Cytoplasm-to-nucleus translocation of a herpesvirus tegument protein during cell division. Journal of virology, 74(5), pp.2131–2141. Elton, D., Simpson-Holley, M., Archer, K., Medcalf, L., Hallam, R., McCauley, J. & Digard, P., 2001. Interaction of the influenza virus nucleoprotein with the cellular CRM1- mediated nuclear export pathway. J Virol, 75(1), pp.408–19. Available at: http://www.ncbi.nlm.nih.gov/htbin- post/Entrez/query?db=m&form=6&dopt=r&uid=11119609\nhttp://jvi.asm.org/cgi/content/full/75/1/408\nhttp://jvi.asm.org/cgi/c ontent/abstract/75/1/408. Enninga, J., Levy, D.E., Blobel, G. & Fontoura, B.M.A., 2002. Role of nucleoporin induction in releasing an mRNA nuclear export block. Science (New York, N.Y.), 295(5559), pp.1523–5. Available at: http://www.sciencemag.org/content/295/5559/1523 [Accessed April 12, 2015]. Epstein, M.A., Hummeler, K. & Berkaloff, A., 1963. THE ENTRY AND DISTRIBUTION OF HERPES VIRUS AND COLLOIDAL GOLD IN HELA CELLS AFTER CONTACT IN SUSPENSION ( From The Bland-Sutton Institu2e of Pathology , Middlesex Hospital Medical School , Landon , England ) ( Received for publication , September 24 , 196. Fadok, V.A., Bratton, D.L. & Henson, P.M., 2001. Phagocyte receptors for apoptotic cells: Recognition, uptake, and consequences. Journal of Clinical Investigation, 108(7), pp.957–962. Fanara, P., Hodel, M.R., Corbett, A.H. & Hodel, A.E., 2000. Quantitative analysis of nuclear localization signal (NLS)-importin ?? interaction through fluorescence depolarization: Evidence for auto-inhibitory regulation of NLS binding. Journal of Biological Chemistry, 275(28),

- 204 -

References

pp.21218–21223. Farnsworth, A., Wisner, T.W. & Johnson, D.C., 2007. Cytoplasmic residues of herpes simplex virus glycoprotein gE required for secondary envelopment and binding of tegument proteins VP22 and UL11 to gE and gD. Journal of virology, 81(1), pp.319–331. Fassati, A., Görlich, D., Harrison, I., Zaytseva, L. & Mingot, J.M., 2003. Nuclear import of HIV-1 intracellular reverse transcription complexes is mediated by importin 7. EMBO Journal, 22(14), pp.3675–3685. Featherstone, C., Darby, M.K. & Gerace, L., 1988. A monoclonal antibody against the nuclear pore complex inhibits nucleocytoplasmic transport of protein and RNA in vivo. The Journal of cell biology, 107(4), pp.1289–1297. Fenwick, M.L. & Clark, J., 1982. Early and delayed shut-off of host protein synthesis in cells infected with herpes simplex virus. Journal of General Virology, 61(1), pp.121–125. Finlay, D.R., Newmeyer, D.D., Price, T.M. & Forbes, D.J., 1987. Inhibition of in vitro nuclear transport by a lectin that binds to nuclear pores. Journal of Cell Biology, 104(2), pp.189–200. Flanagan, W.M., Corthésy, B., Bram, R.J. & Crabtree, G.R., 1991. Nuclear association of a T-cell transcription factor blocked by FK-506 and cyclosporin A. Nature, 352(6338), pp.803–807. Flemington, E.K., 2001. Herpesvirus Lytic Replication and the Cell Cycle : Arresting New Developments MINIREVIEW Herpesvirus Lytic Replication and the Cell Cycle : Arresting New Developments. Journal of virology, 75(10), pp.4475–4481. Available at: http://jvi.asm.org/content/75/10/4475.full.pdf+html. Flint, S.J., 2009. Principles of virology, Third edition. Available at: http://public.eblib.com/choice/publicfullrecord.aspx?p=605163. Foisner, R., 2001. Inner nuclear membrane proteins and the nuclear lamina. Journal of cell science, 114(Pt 21), pp.3791–3792. Fornerod, M., Ohno, M., Yoshida, M. & Mattaj, I.W., 1997. CRM1 is an export receptor for leucine-rich nuclear export signals. Cell, 90(6), pp.1051–60. Available at: http://www.ncbi.nlm.nih.gov/pubmed/9323133 [Accessed March 26, 2015]. Fortes, P., Beloso, A. & Ortín, J., 1994. Influenza virus NS1 protein inhibits pre-mRNA splicing and blocks mRNA nucleocytoplasmic transport. The EMBO journal, 13(3), pp.704–12. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=394862&tool=pmcentrez&rendertype=abstract [Accessed April 12, 2015]. Frenkel, N., Jacob, R.J., Honess, R.W., Hayward, G.S., Locker, H. & Roizman, B., 1975. Anatomy of herpes simplex virus DNA. III. Characterization of defective DNA molecules and biological properties of virus populations containing them. Journal of virology, 16(1), pp.153–167. Fricke, T., Valle-Casuso, J.C., White, T.E., Brandariz-Nuñez, A., Bosche, W.J., Reszka, N., Gorelick, R. & Diaz-Griffero, F., 2013. The ability of TNPO3-depleted cells to inhibit HIV-1 infection requires CPSF6. Retrovirology, 10, p.46. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3695788&tool=pmcentrez&rendertype=abstract. Fuchs, W., Granzow, H., Klupp, B.G., Kopp, M. & Mettenleiter, T.C., 2002. The UL48 tegument protein of pseudorabies virus is critical for intracytoplasmic assembly of infectious virions. Journal of virology, 76(13), pp.6729–6742. Fuchs, W., Klupp, B.G., Granzow, H. & Mettenleiter, T.C., 2004. Essential Function of the Pseudorabies Virus UL36 Gene Product Is Independent of Its Interaction with the UL37 Protein Essential Function of the Pseudorabies Virus UL36 Gene Product Is Independent of Its Interaction with the UL37 Protein. , 78(21), pp.11879–11889. Fuchs, W., Klupp, B.G., Granzow, H., Osterrieder, N. & Mettenleiter, T.C., 2002. The interacting UL31 and UL34 gene products of pseudorabies virus are involved in egress from the host-cell nucleus and represent components of primary enveloped but not mature virions. Journal of virology, 76(1), pp.364–378. Fuller, A.O. & Spear, P.G., 1987. Anti-glycoprotein D antibodies that permit adsorption but block infection by herpes simplex virus 1 prevent virion-cell fusion at the cell surface. Proceedings of the National Academy of Sciences of the United States of America, 84(15), pp.5454– 5458. Galbis-Martínez, M., Saenz, L., Ramírez, P., Parrilla, P. & Yélamos, J., 2010. Poly(ADP-ribose) polymerase-1 modulates interferon-γ-inducible protein (IP)-10 expression in murine embryonic fibroblasts by stabilizing IP-10 mRNA. Molecular Immunology, 47(7-8), pp.1492–1499. Gallay, P., Hope, T., Chin, D. & Trono, D., 1997. HIV-1 infection of nondividing cells through the recognition of integrase by the importin/karyopherin pathway. Proceedings of the National Academy of Sciences of the United States of America, 94(18), pp.9825– 9830. Gallo, S.A., Finnegan, C.M., Viard, M., Raviv, Y., Dimitrov, A., Rawat, S.S., Puri, A., Durell, S. & Blumenthal, R., 2003. The HIV Env-mediated fusion reaction. Biochimica et Biophysica Acta - Biomembranes, 1614(1), pp.36–50. Gartner, S., Markovits, P., Markovitz, D.M., Kaplan, M.H., Gallo, R.C. & Popovic, M., 1986. The role of mononuclear phagocytes in HTLV-III/LAV infection. Science (New York, N.Y.), 233(4760), pp.215–219. Gershey, E.L. & Diacumakos, E.G., 1978. Simian virus 40 production after viral uncoating in the CV-1 cell nucleus. Journal of virology, 28(1), pp.415–416. Gewirtz, D.A., 1999. A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochemical Pharmacology, 57(7), pp.727–741. Ghildyal, R., Jordan, B., Li, D., Dagher, H., Bardin, P.G., Gern, J.E. & Jans, D.A., 2009. Rhinovirus 3C protease can localize in the nucleus and alter active and passive nucleocytoplasmic transport. Journal of virology, 83(14), pp.7349–7352. Gibson, W. & Roizman, B., 1972. Proteins specified by herpes simplex virus. 8. Characterization and composition of multiple capsid forms of subtypes 1 and 2. Journal of virology, 10(5), pp.1044–1052. Gilchrist, D., Mykytka, B. & Rexach, M., 2002. Accelerating the rate of disassembly of karyopherin??cargo complexes. Journal of Biological

- 205 -

References

Chemistry, 277(20), pp.18161–18172. Goldfarb, D.S., Corbett, A.H., Mason, D.A., Harreman, M.T. & Adam, S. a, 2004. Importin alpha: a multipurpose nuclear-transport receptor. Trends in cell biology, 14(9), pp.505–14. Available at: http://www.ncbi.nlm.nih.gov/pubmed/15350979 [Accessed September 12, 2014]. Görlich, D. & Kutay, U., 1999. Transport between the cell nucleus and the cytoplasm. Annual review of cell and developmental biology, 15, pp.607–60. Available at: http://www.ncbi.nlm.nih.gov/pubmed/10611974. Grady, S.L., Hwang, J., Vastag, L., Rabinowitz, J.D. & Shenk, T., 2012. Herpes Simplex Virus 1 Infection Activates Poly(ADP-Ribose) Polymerase and Triggers the Degradation of Poly(ADP-Ribose) Glycohydrolase. Journal of Virology, 86(15), pp.8259–8268. Granzow, H., Klupp, B.G., Fuchs, W., Veits, J., Osterrieder, N. & Mettenleiter, T.C., 2001. Egress of alphaherpesviruses: comparative ultrastructural study. Journal of virology, 75(8), pp.3675–3684. Granzow, H., Klupp, B.G. & Mettenleiter, T.C., 2005. Entry of pseudorabies virus: an immunogold-labeling study. Journal of virology, 79(5), pp.3200–3205. Greber, U.F., Suomalainen, M., Stidwill, R.P., Boucke, K., Ebersold, M.W. & Helenius, a, 1997. The role of the nuclear pore complex in adenovirus DNA entry. The EMBO journal, 16(19), pp.5998–6007. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1170230&tool=pmcentrez&rendertype=abstract. Greber, U.F., Webster, P., Weber, J. & Helenius, A., 1996. The role of the adenovirus protease on virus entry into cells. The EMBO journal, 15(8), pp.1766–1777. Green, M. a, Sviland, L., Malcolm, a J. & Pearson, a D., 1989. Improved method for immunoperoxidase detection of membrane antigens in frozen sections. Journal of clinical pathology, 42(8), pp.875–880. Griffiths, S.J., Koegl, M., Boutell, C., Zenner, H.L., Crump, C.M., Pica, F., Gonzalez, O., Friedel, C.C., Barry, G., Martin, K., Craigon, M.H., Chen, R., Kaza, L.N., Fossum, E., Fazakerley, J.K., Efstathiou, S., Volpi, A., Zimmer, R., Ghazal, P. & Haas, J., 2013. A Systematic Analysis of Host Factors Reveals a Med23-Interferon-λ Regulatory Axis against Herpes Simplex Virus Type 1 Replication. PLoS Pathogens, 9(8). Grimm, K.S., Klupp, B.G., Granzow, H., Muller, F.M., Fuchs, W. & Mettenleiter, T.C., 2012. Analysis of Viral and Cellular Factors Influencing Herpesvirus-Induced Nuclear Envelope Breakdown. Journal of Virology, 86, pp.6512–6521. Grove, J. & Marsh, M., 2011. The cell biology of receptor-mediated virus entry. Journal of Cell Biology, 195(7), pp.1071–1082. Gustin, K.E., 2003. Inhibition of nucleo-cytoplasmic trafficking by RNA viruses: Targeting the nuclear pore complex. Virus Research, 95(1-2), pp.35–44. Gustin, K.E. & Sarnow, P., 2001. Effects of poliovirus infection on nucleo-cytoplasmic trafficking and nuclear pore complex composition. EMBO Journal, 20(1-2), pp.240–249. Gustin, K.E. & Sarnow, P., 2002. Inhibition of nuclear import and alteration of nuclear pore complex composition by rhinovirus. Journal of virology, 76(17), pp.8787–8796. Güttinger, S., Laurell, E. & Kutay, U., 2009. Orchestrating nuclear envelope disassembly and reassembly during mitosis. Nature reviews. Molecular cell biology, 10(3), pp.178–191. Hadigal, S. & Shukla, D., 2013. Exploiting herpes simplex virus entry for novel therapeutics. Viruses, 5(6), pp.1447–1465. Haffar, O.K., Popov, S., Dubrovsky, L., Agostini, I., Tang, H., Pushkarsky, T., Nadler, S.G. & Bukrinsky, M., 2000. Two nuclear localization signals in the HIV-1 matrix protein regulate nuclear import of the HIV-1 pre-integration complex. Journal of molecular biology, 299(2), pp.359–68. Available at: http://www.ncbi.nlm.nih.gov/pubmed/10860744 [Accessed October 30, 2014]. Hagglund, R. & Roizman, B., 2004. Role of ICP0 in the Strategy of Conquest of the Host Cell by Herpes Simplex Virus 1 MINIREVIEW Role of ICP0 in the Strategy of Conquest of the Host Cell by Herpes Simplex Virus 1. , 78(5), pp.2169–2178. Haglund, K. & Dikic, I., 2012. The role of ubiquitylation in receptor endocytosis and endosomal sorting. Journal of Cell Science, 125(2), pp.265– 275. Hansen, J., Qing, K. & Srivastava, A., 2001. Infection of purified nuclei by adeno-associated virus 2. Molecular therapy : the journal of the American Society of Gene Therapy, 4(4), pp.289–296. Harel, A. & Forbes, D.J., 2004. Importin beta: Conducting a much larger cellular symphony. Molecular Cell, 16(3), pp.319–330. Harland, J., Dunn, P., Cameron, E., Conner, J. & Brown, S.M., 2003. The herpes simplex virus (HSV) protein ICP34.5 is a virion component that forms a DNA-binding complex with proliferating cell nuclear antigen and HSV replication proteins. Journal of neurovirology, 9(4), pp.477–488. Harley, C. a, Dasgupta, a & Wilson, D.W., 2001. Characterization of herpes simplex virus-containing organelles by subcellular fractionation: role for organelle acidification in assembly of infectious particles. Journal of virology, 75(3), pp.1236–1251. Harreman, M.T., Hodel, M.R., Fanara, P., Hodel, A.E. & Corbett, A.H., 2003. The auto-inhibitory function of importin ?? is essential in vivo. Journal of Biological Chemistry, 278(8), pp.5854–5863. Hasebe, R., Sasaki, M., Sawa, H., Wada, R., Umemura, T. & Kimura, T., 2009. Infectious entry of equine herpesvirus-1 into host cells through different endocytic pathways. Virology, 393(2), pp.198–209. Available at: http://dx.doi.org/10.1016/j.virol.2009.07.032. Hefferon, K.L., Oomens, a G., Monsma, S. a, Finnerty, C.M. & Blissard, G.W., 1999. Host cell receptor binding by baculovirus GP64 and kinetics of virion entry. Virology, 258(2), pp.455–468. Heine, J.W., Honess, R.W., Cassai, E. & Roizman, B., 1974. Proteins specified by herpes simplex virus. XII. The virion polypeptides of type 1 strains. Journal of virology, 14(3), pp.640–651. Henaff, D., Rémillard-Labrosse, G., Loret, S. & Lippé, R., 2013. Analysis of the early steps of herpes simplex virus 1 capsid tegumentation. Journal of virology, 87(9), pp.4895–906. Available at:

- 206 -

References

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3624315&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Hennig, T., Abaitua, F. & O’Hare, P., 2014. Functional analysis of nuclear localisation signals in VP1-2 homologues from all herpesvirus subfamilies. Journal of virology, 44(February), pp.5391–405. Her, L., 1997. Inhibition of Ran Guanosine Triphosphatase-Dependent Nuclear Transport by the Matrix Protein of Vesicular Stomatitis Virus. Science, 276(5320), pp.1845–1848. Available at: http://www.sciencemag.org/content/276/5320/1845 [Accessed April 12, 2015]. Herold, B.C., WuDunn, D., Soltys, N. & Spear, P.G., 1991. Glycoprotein C of herpes simplex virus type 1 plays a principal role in the adsorption of virus to cells and in infectivity. Journal of virology, 65(3), pp.1090–1098. Hindley, C.E., Lawrence, F.J. & Matthews, D. a, 2007. A role for transportin in the nuclear import of adenovirus core proteins and DNA. Traffic (Copenhagen, Denmark), 8(10), pp.1313–22. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2171040&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Hodel, A.E., Harreman, M.T., Pulliam, K.F., Harben, M.E., Holmes, J.S., Hodel, M.R., Berland, K.M. & Corbett, A.H., 2006. Nuclear localization signal receptor affinity correlates with in vivo localization in Saccharomyces cerevisiae. Journal of Biological Chemistry, 281(33), pp.23545–23556. Hofemeister, H. & O’Hare, P., 2008. Nuclear pore composition and gating in herpes simplex virus-infected cells. Journal of virology, 82(17), pp.8392–8399. Hogle, J.M., 2002. Poliovirus cell entry: common structural themes in viral cell entry pathways. Annual review of microbiology, 56, pp.677– 702. Holleman, J. & Marchese, A., 2014. The ubiquitin ligase deltex-3l regulates endosomal sorting of the G protein-coupled receptor CXCR4. Molecular biology of the cell, 25(12), pp.1892–904. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24790097. Hollinshead, M., Johns, H.L., Sayers, C.L., Gonzalez-Lopez, C., Smith, G.L. & Elliott, G., 2012. Endocytic tubules regulated by Rab GTPases 5 and 11 are used for envelopment of herpes simplex virus. The EMBO Journal. Honess, R.W. & Roizman, B., 1973. Proteins Specified by Herpes Simplex Virus XI. Identification and Relative Molar Rates of Synthesis of Structural and Nonstructural Herpes Virus Polypeptides in the Infected Cell. Journal of virology, 12(6), pp.1347–1365. Honess, R.W. & Roizman, B., 1974. Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. Journal of virology, 14(1), pp.8–19. Horan, K.A., Hansen, K., Jakobsen, M.R., Holm, C.K., Søby, S., Unterholzner, L., Thompson, M., West, J.A., Iversen, M.B., Rasmussen, S.B., Ellermann-Eriksen, S., Kurt-Jones, E., Landolfo, S., Damania, B., Melchjorsen, J., Bowie, A.G., Fitzgerald, K.A. & Paludan, S.R., 2013. Proteasomal degradation of herpes simplex virus capsids in macrophages releases DNA to the cytosol for recognition by DNA sensors. Journal of immunology (Baltimore, Md. : 1950), 190(5), pp.2311–9. Available at: http://www.jimmunol.org/content/190/5/2311.full [Accessed April 28, 2015]. Hoyt, C.C., Bouchard, R.J. & Tyler, K.L., 2004. Novel Nuclear Herniations Induced by Nuclear Localization of a Viral Protein. Society, 78(12), pp.6360–6369. Https://www.fli.bund.de/typo3temp/pics/201d5a0ec0.jpg, The Herpesvirus Life Cycle. Available at: https://www.fli.bund.de/typo3temp/pics/201d5a0ec0.jpg [Accessed May 17, 2015]. Huang, D.W., Sherman, B.T. & Lempicki, R.A., 2009. Bioinformatics enrichment tools: Paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Research, 37(1), pp.1–13. Huang, D.W., Sherman, B.T. & Lempicki, R.A., 2009. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature protocols, 4(1), pp.44–57. Huang, D.W., Sherman, B.T., Tan, Q., Collins, J.R., Alvord, W.G., Roayaei, J., Stephens, R., Baseler, M.W., Lane, H.C. & Lempicki, R.A., 2007. The DAVID Gene Functional Classification Tool: a novel biological module-centric algorithm to functionally analyze large gene lists. Genome biology, 8(9), p.R183. Huang, S., Chen, J., Chen, Q., Wang, H., Yao, Y., Chen, J. & Chen, Z., 2013. A second CRM1-dependent nuclear export signal in the influenza A virus NS2 protein contributes to the nuclear export of viral ribonucleoproteins. Journal of virology, 87(2), pp.767–78. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3554077&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Hübner, S., Xiao, C.Y. & Jans, D.A., 1997. The protein kinase CK2 site (Ser(111/112)) enhances recognition of the simian virus 40 large T- antigen nuclear localization sequence by importin. Journal of Biological Chemistry, 272(27), pp.17191–17195. Huen, M.S.Y. & Chen, J., 2008. The DNA damage response pathways: at the crossroad of protein modifications. Cell research, 18(1), pp.8–16. De Iaco, A., Santoni, F., Vannier, A., Guipponi, M., Antonarakis, S. & Luban, J., 2013. TNPO3 protects HIV-1 replication from CPSF6-mediated capsid stabilization in the host cell cytoplasm. Retrovirology, 10(1), p.20. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3599327&tool=pmcentrez&rendertype=abstract. Ishikawa, H., Ma, Z. & Barber, G.N., 2009. STING regulates intracellular DNA-mediated, type I interferon-dependent innate immunity. Nature, 461(7265), pp.788–792. Available at: http://dx.doi.org/10.1038/nature08476. Izumiya, Y., Jang, H.K., Sugawara, M., Ikeda, Y., Miura, R., Nishimura, Y., Nakamura, K., Miyazawa, T., Kai, C. & Mikami, T., 1999. Identification and transcriptional analysis of the homologues of the herpes simplex virus type 1 UL30 to UL40 genes in the genome of nononcogenic Marek’s disease virus serotype 2. Journal of General Virology, 80(9), pp.2417–2422. Jackson, S.P. & Durocher, D., 2013. Regulation of DNA Damage Responses by Ubiquitin and SUMO. Molecular Cell, 49(5), pp.795–807.

- 207 -

References

Available at: http://dx.doi.org/10.1016/j.molcel.2013.01.017. Jacquot, G., Le Rouzic, E., David, A., Mazzolini, J., Bouchet, J., Bouaziz, S., Niedergang, F., Pancino, G. & Benichou, S., 2007. Localization of HIV-1 Vpr to the nuclear envelope: impact on Vpr functions and virus replication in macrophages. Retrovirology, 4, p.84. Jain, J., McCaffrey, P.G., Miner, Z., Kerppola, T.K., Lambert, J.N., Verdine, G.L., Curran, T. & Rao, A., 1993. The T-cell transcription factor NFATp is a substrate for calcineurin and interacts with Fos and Jun. Nature, 365(6444), pp.352–355. Jäkel, S., Albig, W., Kutay, U., Bischoff, F.R., Schwamborn, K., Doenecke, D. & Görlich, D., 1999. The importin β/importin 7 heterodimer is a functional nuclear import receptor for histone H1. EMBO Journal, 18(9), pp.2411–2423. Jambunathan, N., Chouljenko, D., Desai, P., Charles, A.-S., Subramanian, R., Chouljenko, V.N. & Kousoulas, K.G., 2014. Herpes Simplex Virus 1 Protein UL37 Interacts with Viral Glycoprotein gK and Membrane Protein UL20 and Functions in Cytoplasmic Virion Envelopment. Journal of virology, 88(11), pp.5927–35. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24600000. Jerome, K.R., Chen, Z., Lang, R., Torres, M.R., Hofmeister, J., Smith, S., Fox, R., Froelich, C.J. & Corey, L., 2001. HSV and glycoprotein J inhibit caspase activation and apoptosis induced by granzyme B or Fas. Journal of immunology (Baltimore, Md. : 1950), 167(7), pp.3928– 3935. Jerome, K.R., Fox, R., Chen, Z., Sears, A.E., Lee, H. y & Corey, L., 1999. Herpes simplex virus inhibits apoptosis through the action of two genes, Us5 and Us3. Journal of virology, 73(11), pp.8950–8957. Johnson, L.A., Li, L. & Sandri-Goldin, R.M., 2009. The Cellular RNA Export Receptor TAP/NXF1 Is Required for ICP27-Mediated Export of Herpes Simplex Virus 1 RNA, but the TREX Complex Adaptor Protein Aly/REF Appears To Be Dispensable. Journal of Virology, 83(13), pp.6335– 6346. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2698537&tool=pmcentrez&rendertype=abstract [Accessed April 12, 2015]. Jones, P.C. & Roizman, B., 1979. Regulation of herpesvirus macromolecular synthesis. VIII. The transcription program consists of three phases during which both extent of transcription and accumulation of RNA in the cytoplasm are regulated. Journal of virology, 31(2), pp.299– 314. Jovasevic, V., Liang, L. & Roizman, B., 2008. Proteolytic cleavage of VP1-2 is required for release of herpes simplex virus 1 DNA into the nucleus. Journal of virology, 82(7), pp.3311–3319. Juszczynski, P., Kutok, J.L., Li, C., Mitra, J., Aguiar, R.C.T. & Shipp, M.A., 2006. BAL1 and BBAP are regulated by a gamma interferon-responsive bidirectional promoter and are overexpressed in diffuse large B-cell lymphomas with a prominent inflammatory infiltrate. Molecular and cellular biology, 26(14), pp.5348–5359. Kalab, P. & Heald, R., 2008. The RanGTP gradient - a GPS for the mitotic spindle. Journal of cell science, 121(Pt 10), pp.1577–1586. Kalderon, D., Roberts, B.L., Richardson, W.D. & Smith, A.E., 1984. A short amino acid sequence able to specify nuclear location. Cell, 39(3 Pt 2), pp.499–509. Kalverda, B., Pickersgill, H., Shloma, V. V. & Fornerod, M., 2010. Nucleoporins Directly Stimulate Expression of Developmental and Cell-Cycle Genes Inside the Nucleoplasm. Cell, 140(3), pp.360–371. Available at: http://dx.doi.org/10.1016/j.cell.2010.01.011. Kann, M., Sodeik, B., Vlachou, a, Gerlich, W.H. & Helenius, a, 1999. Phosphorylation-dependent binding of hepatitis B virus core particles to the nuclear pore complex. The Journal of cell biology, 145(1), pp.45–55. Kappel, C., Zachariae, U., Dölker, N. & Grubmüller, H., 2010. An unusual hydrophobic core confers extreme flexibility to HEAT repeat proteins. Biophysical Journal, 99(5), pp.1596–1603. Karlas, A., Machuy, N., Shin, Y., Pleissner, K.-P., Artarini, A., Heuer, D., Becker, D., Khalil, H., Ogilvie, L. a, Hess, S., Mäurer, A.P., Müller, E., Wolff, T., Rudel, T. & Meyer, T.F., 2010. Genome-wide RNAi screen identifies human host factors crucial for influenza virus replication. Nature, 463(7282), pp.818–822. Kato, A., Liu, Z., Minowa, A., Imai, T., Tanaka, M., Sugimoto, K., Nishiyama, Y., Arii, J. & Kawaguchi, Y., 2011. Herpes Simplex Virus 1 Protein Kinase Us3 and Major Tegument Protein UL47 Reciprocally Regulate Their Subcellular Localization in Infected Cells. Journal of virology, 85(18), pp.9599–9613. Kattenhorn, L.M., Korbel, G. a., Kessler, B.M., Spooner, E. & Ploegh, H.L., 2005. A deubiquitinating enzyme encoded by HSV-1 belongs to a family of cysteine proteases that is conserved across the family Herpesviridae. Molecular Cell, 19, pp.547–557. Kehat, I., Accornero, F., Aronow, B.J. & Molkentin, J.D., 2011. Modulation of chromatin position and gene expression by HDAC4 interaction with nucleoporins. Journal of Cell Biology, 193(1), pp.21–29. Kelly, B.J., Fraefel, C., Cunningham, A.L. & Diefenbach, R.J., 2009. Functional roles of the tegument proteins of herpes simplex virus type 1. Virus Research, 145, pp.173–186. Keminer, O. & Peters, R., 1999. Permeability of single nuclear pores. Biophysical journal, 77(1), pp.217–228. Kim, E., Lee, Y., Lee, H.J., Kim, J.S., Song, B.S., Huh, J.W., Lee, S.R., Kim, S.U., Kim, S.H., Hong, Y., Shim, I. & Chang, K.T., 2010. Implication of mouse Vps26b-Vps29-Vps35 retromer complex in sortilin trafficking. Biochemical and Biophysical Research Communications, 403(2), pp.167–171. Kingston, D., Chang, H., Ensser, A., Lee, H.-R., Lee, J., Lee, S.-H., Jung, J.U. & Cho, N.-H., 2011. Inhibition of Retromer Activity by Herpesvirus Saimiri Tip Leads to CD4 Downregulation and Efficient T Cell Transformation. Journal of Virology, 85(20), pp.10627–10638. Klupp, B.G., Fuchs, W., Granzow, H., Nixdorf, R. & Mettenleiter, T.C., 2002. Pseudorabies virus UL36 tegument protein physically interacts with the UL37 protein. Journal of virology, 76(6), pp.3065–3071. Klupp, B.G., Granzow, H., Fuchs, W., Keil, G.M., Finke, S. & Mettenleiter, T.C., 2007. Vesicle formation from the nuclear membrane is induced by coexpression of two conserved herpesvirus proteins. Proceedings of the National Academy of Sciences of the United States of America, 104(17), pp.7241–7246.

- 208 -

References

Klupp, B.G., Granzow, H. & Mettenleiter, T.C., 2011. Nuclear envelope breakdown can substitute for primary envelopment-mediated nuclear Egress of herpesviruses. Journal of virology, 85(16), pp.8285–8292. Klupp, B.G., Granzow, H., Mundt, E. & Thomas, C., 2001. Pseudorabies Virus UL37 Gene Product Is Involved in Secondary Envelopment Pseudorabies Virus UL37 Gene Product Is Involved in Secondary Envelopment. , 75(19), pp.8927–8936. Knipe, D.M., Batterson, W., Nosal, C., Roizman, B. & Buchan, a, 1981. Molecular genetics of herpes simplex virus. VI. Characterization of a temperature-sensitive mutant defective in the expression of all early viral gene products. Journal of virology, 38(2), pp.539–547. Knipe, D.M. & Howley, P.M., 2013. Fields Virology 6th edition; Chapter 41: Orthomyxoviruses 6th ed., Ko, D.H., Cunningham, A.L. & Diefenbach, R.J., 2010. The major determinant for addition of tegument protein pUL48 (VP16) to capsids in herpes simplex virus type 1 is the presence of the major tegument protein pUL36 (VP1/2). Journal of virology, 84(3), pp.1397–1405. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2812353&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. von Kobbe, C., van Deursen, J.M.A., Rodrigues, J.P., Sitterlin, D., Bachi, A., Wu, X., Wilm, M., Carmo-Fonseca, M. & Izaurralde, E., 2000. Vesicular Stomatitis Virus Matrix Protein Inhibits Host Cell Gene Expression by Targeting the Nucleoporin Nup98. Molecular Cell, 6(5), pp.1243–1252. Available at: http://www.sciencedirect.com/science/article/pii/S1097276500001209 [Accessed April 12, 2015]. Koehler, M., Fiebeler, A., Hartwig, M., Thiel, S., Prehn, S., Kettritz, R., Luft, F. & Hartmann, E., 2002. Differential expression of classical nuclear transport factors during cellular proliferation and differentiation. Cellular Physiology and Biochemistry, 12(5-6), pp.335–344. König, R., Stertz, S., Zhou, Y., Inoue, A., Hoffmann, H.-H., Bhattacharyya, S., Alamares, J.G., Tscherne, D.M., Ortigoza, M.B., Liang, Y., Gao, Q., Andrews, S.E., Bandyopadhyay, S., De Jesus, P., Tu, B.P., Pache, L., Shih, C., Orth, A., Bonamy, G., Miraglia, L., Ideker, T., García-Sastre, A., Young, J. a T., Palese, P., Shaw, M.L. & Chanda, S.K., 2010. Human host factors required for influenza virus replication. Nature, 463(7282), pp.813–817. Kopp, M., Klupp, B.G., Granzow, H., Fuchs, W. & Mettenleiter, T.C., 2002. Identification and characterization of the pseudorabies virus tegument proteins UL46 and UL47: role for UL47 in virion morphogenesis in the cytoplasm. Journal of virology, 76(17), pp.8820–8833. Kosugi, S., Hasebe, M., Entani, T., Takayama, S., Tomita, M. & Yanagawa, H., 2008. Design of Peptide Inhibitors for the Importin α/β Nuclear Import Pathway by Activity-Based Profiling. Chemistry and Biology, 15(9), pp.940–949. Kosugi, S., Hasebe, M., Matsumura, N., Takashima, H., Miyamoto-Sato, E., Tomita, M. & Yanagawa, H., 2009. Six classes of nuclear localization signals specific to different binding grooves of importin alpha. Journal of Biological Chemistry, 284(1), pp.478–485. Available at: http://www.ncbi.nlm.nih.gov/pubmed/19001369 [Accessed October 3, 2014]. Kosugi, S., Hasebe, M., Tomita, M. & Yanagawa, H., 2009. Systematic identification of cell cycle-dependent yeast nucleocytoplasmic shuttling proteins by prediction of composite motifs. Proceedings of the National Academy of Sciences of the United States of America, 106(25), pp.10171–10176. Krautwald, M., Fuchs, W., Klupp, B.G. & Mettenleiter, T.C., 2009. Translocation of incoming pseudorabies virus capsids to the cell nucleus is delayed in the absence of tegument protein pUL37. Journal of virology, 83(7), pp.3389–3396. Krawczyk, E., Hanover, J.A., Schlegel, R. & Suprynowicz, F.A., 2008. Karyopherin β3: A new cellular target for the HPV-16 E5 oncoprotein. Biochemical and Biophysical Research Communications, 371(4), pp.684–688. Krenciute, G., Liu, S., Yucer, N., Shi, Y., Ortiz, P., Liu, Q., Kim, B.J., Odejimi, A.O., Leng, M., Qin, J. & Wang, Y., 2013. Nuclear BAG6-UBL4A-GET4 complex mediates DNA damage signaling and cell death. Journal of Biological Chemistry, 288(28), pp.20547–20557. Krishnan, L., Matreyek, K.A., Oztop, I., Lee, K., Tipper, C.H., Li, X., Dar, M.J., Kewalramani, V.N. & Engelman, A., 2010. The requirement for cellular transportin 3 (TNPO3 or TRN-SR2) during infection maps to human immunodeficiency virus type 1 capsid and not integrase. Journal of virology, 84(1), pp.397–406. Kristensson, K., Lycke, E., Roytta, M., Svennerholm, B. & Vahlne, a., 1986. Neuritic transport of herpes simplex virus in rat sensory neurons in vitro. Effects of substances interacting with microtubular function and axonal flow [nocodazole, taxol and erythro-9-3-(2- hydroxynonyl)adenine]. Journal of General Virology, 67(9), pp.2023–2028. Krummenacher, C., Nicola, A. V, Whitbeck, J.C., Lou, H., Hou, W., Lambris, J.D., Geraghty, R.J., Spear, P.G., Cohen, G.H. & Eisenberg, R.J., 1998. Herpes simplex virus glycoprotein D can bind to poliovirus receptor-related protein 1 or herpesvirus entry mediator, two structurally unrelated mediators of virus entry. Journal of virology, 72(9), pp.7064–7074. Kuhn, J., Leege, T., Klupp, B.G., Granzow, H., Fuchs, W. & Mettenleiter, T.C., 2008. Partial functional complementation of a pseudorabies virus UL25 deletion mutant by herpes simplex virus type 1 pUL25 indicates overlapping functions of alphaherpesvirus pUL25 proteins. Journal of virology, 82(12), pp.5725–5734. Kurane, I., Tsuchiya, Y., Sekizawa, T. & Kumagai, K., 1984. Inhibition by indomethacin of in vitro reactivation of latent herpes simplex virus type 1 in murine trigeminal ganglia. Journal of General Virology, 65(10), pp.1665–1674. Kutay, U., Izaurralde, E., Bischoff, F.R., Mattaj, I.W. & Görlich, D., 1997. Dominant-negative mutants of importin-beta block multiple pathways of import and export through the nuclear pore complex. The EMBO journal, 16(6), pp.1153–1163. Lange, A., McLane, L.M., Mills, R.E., Devine, S.E. & Corbett, A.H., 2010. Expanding the definition of the classical bipartite nuclear localization signal. Traffic, 11(3), pp.311–323. Lange, A., Mills, R.E., Lange, C.J., Stewart, M., Devine, S.E. & Corbett, A.H., 2007. Classical nuclear localization signals: Definition, function, and interaction with importin α. Journal of Biological Chemistry, 282(8), pp.5101–5105. Leach, N., Bjerke, S.L., Christensen, D.K., Bouchard, J.M., Mou, F., Park, R., Baines, J., Haraguchi, T. & Roller, R.J., 2007. Emerin is hyperphosphorylated and redistributed in herpes simplex virus type 1-infected cells in a manner dependent on both UL34 and US3. Journal of virology, 81(19), pp.10792–803. Available at:

- 209 -

References

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2045475&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Lee, J.I.-H., Luxton, G.W.G. & Smith, G.A., 2006. Identification of an essential domain in the herpesvirus VP1/2 tegument protein: the carboxy terminus directs incorporation into capsid assemblons. Journal of virology, 80(24), pp.12086–12094. Lee, K., Ambrose, Z., Martin, T.D., Oztop, I., Mulky, A., Julias, J.G., Vandegraaff, N., Baumann, J.G., Wang, R., Yuen, W., Takemura, T., Shelton, K., Taniuchi, I., Li, Y., Sodroski, J., Littman, D.R., Coffin, J.M., Hughes, S.H., Unutmaz, D., Engelman, A. & KewalRamani, V.N., 2010. Flexible Use of Nuclear Import Pathways by HIV-1. Cell Host and Microbe, 7(3), pp.221–233. Lee, M.Y.W.T., Zhang, S., Hua, S., Lin, S., Wang, X., Darzynkiewicz, Z., Zhang, Z. & Lee, E.Y.C., 2014. The tail that wags the dog. , pp.23–31. Lee, S.J., Matsuura, Y., Liu, S.M. & Stewart, M., 2005. Structural basis for nuclear import complex dissociation by RanGTP. Nature, 435(7042), pp.693–696. Leege, T., Fuchs, W., Granzow, H., Kopp, M., Klupp, B.G. & Mettenleiter, T.C., 2009. Effects of simultaneous deletion of pUL11 and glycoprotein M on virion maturation of herpes simplex virus type 1. Journal of virology, 83(2), pp.896–907. Leege, T., Granzow, H., Fuchs, W., Klupp, B.G. & Mettenleiter, T.C., 2009. Phenotypic similarities and differences between UL37-deleted pseudorabies virus and herpes simplex virus type 1. The Journal of general virology, 90, pp.1560–1568. Leelawong, M., Lee, J.I. & Smith, G.A., 2012. Nuclear Egress of Pseudorabies Virus Capsids Is Enhanced by a Subspecies of the Large Tegument Protein That Is Lost upon Cytoplasmic Maturation. Journal of Virology, 86(11), pp.6303–6314. Leung, S.W., Harreman, M.T., Hodel, M.R., Hodel, A.E. & Corbett, A.H., 2003. Dissection of the karyopherin ?? nuclear localization signal (NLS)-binding groove: Functional requirements for NLS binding. Journal of Biological Chemistry, 278(43), pp.41947–41953. Leuzinger, H., Ziegler, U., Schraner, E.M., Fraefel, C., Glauser, D.L., Heid, I., Ackermann, M., Mueller, M. & Wild, P., 2005. Herpes simplex virus 1 envelopment follows two diverse pathways. Journal of virology, 79(20), pp.13047–13059. Levin, A., Hayouka, Z., Friedler, A. & Loyter, A., 2010. Transportin 3 and importin α are required for effective nuclear import of HIV-1 integrase in virus-infected cells. , 1(5), pp.422–431. Li, Y., Zhao, L., Wang, S., Xing, J. & Zheng, C., 2012. Identification of a novel NLS of herpes simplex virus type 1 (HSV-1) VP19C and its nuclear localization is required for efficient production of HSV-1. Journal of General Virology, 93, pp.1869–1875. Lilley, C.E., Carson, C.T., Muotri, A.R., Gage, F.H. & Weitzman, M.D., 2005. DNA repair proteins affect the lifecycle of herpes simplex virus 1. Proceedings of the National Academy of Sciences of the United States of America, 102(16), pp.5844–5849. Lilley, C.E., Chaurushiya, M.S., Boutell, C., Everett, R.D. & Weitzman, M.D., 2011. The intrinsic antiviral defense to incoming HSV-1 genomes includes specific DNA repair proteins and is counteracted by the viral protein ICP0. PLoS Pathogens, 7(6). Liptak, L.M., Uprichard, S.L. & Knipe, D.M., 1996. Functional order of assembly of herpes simplex virus DNA replication proteins into prereplicative site structures. Journal of virology, 70(3), pp.1759–1767. Liu, Y., Gong, W., Huang, C.C., Herr, W. & Cheng, X., 1999. Crystal structure of the conserved core of the herpes simplex virus transcriptional regulatory protein VP16. Genes and Development, 13(13), pp.1692–1703. Liu, Y., Li, S. & Wang, Z., 2014. The Role of Cyclooxygenase in Multiplication and Reactivation of HSV-1 in Vestibular Ganglion Neurons. , 2014. Lodish, H.F., Berk, A., Zipursky, S.L., Matsudaira, P., Baltimore, D. & James, D., 2008. Molecular Cell Biology, Available at: http://www.amazon.com/Molecular-Cell-Biology-Harvey-Lodish/dp/0716743663. Long, G., Pan, X., Kormelink, R. & Vlak, J.M., 2006. Functional entry of baculovirus into insect and mammalian cells is dependent on clathrin- mediated endocytosis. Journal of virology, 80(17), pp.8830–8833. Loo, N. Van, Fortunati, E., Ehlert, E., Grosveld, F. & Scholte, B.J., 2001. Baculovirus Infection of Nondividing Mammalian Cells : Mechanisms of Entry and Nuclear Transport of Capsids Baculovirus Infection of Nondividing Mammalian Cells : Mechanisms of Entry and Nuclear Transport of Capsids. , 75(2), pp.961–970. van Loo, N.D., Fortunati, E., Ehlert, E., Rabelink, M., Grosveld, F. & Scholte, B.J., 2001. Baculovirus infection of nondividing mammalian cells: mechanisms of entry and nuclear transport of capsids. Journal of virology, 75(2), pp.961–970. Loret, S., Guay, G. & Lippé, R., 2008. Comprehensive characterization of extracellular herpes simplex virus type 1 virions. Journal of virology, 82(17), pp.8605–8618. Lott, K. & Cingolani, G., 2011. The importin β binding domain as a master regulator of nucleocytoplasmic transport. Biochimica et biophysica acta, 1813(9), pp.1578–92. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3037977&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Lowe, A.R., Tang, J.H., Yassif, J., Graf, M., Huang, W.Y.C., Groves, J.T., Weis, K. & Liphardt, J.T., 2015. Importin-β modulates the permeability of the nuclear pore complex in a Ran-dependent manner. eLife, 17(1), pp.151–152. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4375889&tool=pmcentrez&rendertype=abstract [Accessed March 13, 2015]. Ludwig, S., Wolff, T., Ehrhardt, C., Wurzer, W.J., Reinhardt, J., Planz, O. & Pleschka, S., 2004. MEK inhibition impairs influenza B virus propagation without emergence of resistant variants. FEBS letters, 561(1-3), pp.37–43. Available at: http://www.ncbi.nlm.nih.gov/pubmed/15013748 [Accessed March 30, 2015]. Luxton, G.W.G., Haverlock, S., Coller, K.E., Antinone, S.E., Pincetic, A. & Smith, G.A., 2005. Targeting of herpesvirus capsid transport in axons is coupled to association with specific sets of tegument proteins. Proceedings of the National Academy of Sciences of the United States of America, 102(16), pp.5832–5837. Luxton, G.W.G., Lee, J.I., Schober, J.M., Smith, A., Haverlock-moyns, S. & Smith, G.A., 2006. The Pseudorabies Virus VP1 / 2 Tegument Protein

- 210 -

References

Is Required for Intracellular Capsid Transport The Pseudorabies Virus VP1 / 2 Tegument Protein Is Required for Intracellular Capsid Transport †. Journal of virology, 80(1), pp.201–209. Lyman, M.G. & Enquist, L.W., 2009. Herpesvirus interactions with the host cytoskeleton. Journal of virology, 83(5), pp.2058–2066. Maertens, G.N., Hare, S. & Cherepanov, P., 2010. The mechanism of retroviral integration from X-ray structures of its key intermediates. Nature, 468(7321), pp.326–329. Mahajan, R., Delphin, C., Guan, T., Gerace, L. & Melchior, F., 1997. A small ubiquitin-related polypeptide involved in targeting RanGAP1 to nuclear pore complex protein RanBP2. Cell, 88(1), pp.97–107. Maier, O., Marvin, S.A., Wodrich, H., Campbell, E.M. & Wiethoff, C.M., 2012. Spatiotemporal Dynamics of Adenovirus Membrane Rupture and Endosomal Escape. Journal of Virology, 86(19), pp.10821–10828. Malik, P., Tabarraei, A., Kehlenbach, R.H., Korfali, N., Iwasawa, R., Graham, S. V. & Schirmer, E.C., 2012. Herpes simplex virus ICP27 protein directly interacts with the nuclear pore complex through Nup62, inhibiting host nucleocytoplasmic transport pathways. Journal of Biological Chemistry, 287(15), pp.12277–12292. Marfori, M., Mynott, A., Ellis, J.J., Mehdi, A.M., Saunders, N.F.W., Curmi, P.M., Forwood, J.K., Bodén, M. & Kobe, B., 2011. Molecular basis for specificity of nuclear import and prediction of nuclear localization. Biochimica et biophysica acta, 1813(9), pp.1562–77. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20977914 [Accessed October 30, 2014]. Maric, M., Haugo, A.C., Dauer, W., Johnson, D. & Roller, R.J., 2014. Nuclear envelope breakdown induced by herpes simplex virus type 1 involves the activity of viral fusion proteins. Virology, 460-461, pp.128–137. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0042682214002207. Marjuki, H., Alam, M.I., Ehrhardt, C., Wagner, R., Planz, O., Klenk, H.-D., Ludwig, S. & Pleschka, S., 2006. Membrane accumulation of influenza A virus hemagglutinin triggers nuclear export of the viral genome via protein kinase Calpha-mediated activation of ERK signaling. The Journal of biological chemistry, 281(24), pp.16707–15. Available at: http://www.ncbi.nlm.nih.gov/pubmed/16608852 [Accessed May 10, 2015]. Marschall, M., Marzi, A., aus dem Siepen, P., Jochmann, R., Kalmer, M., Auerochs, S., Lischka, P., Leis, M. & Stamminger, T., 2005. Cellular p32 recruits cytomegalovirus kinase pUL97 to redistribute the nuclear lamina. The Journal of biological chemistry, 280(39), pp.33357– 67. Available at: http://www.jbc.org/content/280/39/33357.abstract?ijkey=10f818950c3a264c6fa97136e38fe3bb68ce9a44&keytype2=tf_ipsecsha [Accessed May 10, 2015]. Martin, K. & Helenius, A., 1991. Transport of incoming influenza virus nucleocapsids into the nucleus. Journal of virology, 65(1), pp.232–244. Mateo, M., Reid, S.P., Leung, L.W., Basler, C.F. & Volchkov, V.E., 2010. Ebolavirus VP24 binding to karyopherins is required for inhibition of interferon signaling. Journal of virology, 84(2), pp.1169–75. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2798383&tool=pmcentrez&rendertype=abstract [Accessed February 27, 2015]. Matlin, K.S., Reggio, H., Helenius, a & Simons, K., 1981. Infectious Entry Pathway of Influenza-Virus in A Canine Kidney-Cell Line. Journal of Cell Biology, 91(3), pp.601–613. Available at: ISI:A1981MT87600001. Matreyek, K. a. & Engelman, A., 2013. Viral and cellular requirements for the nuclear entry of retroviral preintegration nucleoprotein complexes. Viruses, 5(10), pp.2483–2511. Matsuura, Y., Lange, A., Harreman, M.T., Corbett, A.H. & Stewart, M., 2003. Structural basis for Nup2p function in cargo release and karyopherin recycling in nuclear import. EMBO Journal, 22(20), pp.5358–5369. Mboko, W.P., Mounce, B.C., Wood, B.M., Kulinski, J.M., Corbett, J.A. & Tarakanova, V.L., 2012. Coordinate Regulation of DNA Damage and Type I Interferon Responses Imposes an Antiviral State That Attenuates Mouse Gammaherpesvirus Type 68 Replication in Primary Macrophages. Journal of Virology, 86(12), pp.6899–6912. McElwee, M., Beilstein, F., Labetoulle, M., Rixon, F.J. & Pasdeloup, D., 2013. Dystonin/BPAG1 promotes plus-end-directed transport of herpes simplex virus 1 capsids on microtubules during entry. Journal of virology, 87(20), pp.11008–18. Available at: http://www.ncbi.nlm.nih.gov/pubmed/23903849. McGeoch, D.J., Dalrymple, M.A., Davison, A.J., Dolan, A., Frame, M.C., McNab, D., Perry, L.J., Scott, J.E. & Taylor, P., 1988. The complete DNA sequence of the long unique region in the genome of herpes simplx virus type 1. Journal of General Virology, 69(7), pp.1531–1574. Mclane, L.M., Pulliam, K.F., Devine, S.E. & Corbett, A.H., 2008. The Ty1 integrase protein can exploit the classical nuclear protein import machinery for entry into the nucleus. Nucleic Acids Research, 36(13), pp.4317–4326. McNabb, D.S. & Courtney, R.J., 1992. Analysis of the UL36 open reading frame encoding the large tegument protein (ICP1/2) of herpes simplex virus type 1. Journal of virology, 66(12), pp.7581–7584. McNabb, D.S. & Courtney, R.J., 1992. Characterization of the large tegument protein (ICP1/2) of herpes simplex virus type 1. Virology, 190(1), pp.221–232. Meggio, F. & Pinna, L. a, 2003. One-thousand-and-one substrates of protein kinase CK2? The FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 17(3), pp.349–368. Melen, K., Fagerlund, R., Franke, J., Kohler, M., Kinnunen, L. & Julkunen, I., 2003. Importin alpha nuclear localization signal binding sites for STAT1, STAT2, and influenza A virus nucleoprotein. The Journal of biological chemistry, 278(30), pp.28193–28200. Mercer, J. & Helenius, A., 2009. Virus entry by macropinocytosis. Nature cell biology, 11(5), pp.510–520. Mettenleiter, T.C., 2006. Intriguing interplay between viral proteins during herpesvirus assembly or: The herpesvirus assembly puzzle. Veterinary Microbiology, 113(3-4 SPEC. ISS.), pp.163–169.

- 211 -

References

Mettenleiter, T.C., Müller, F., Granzow, H. & Klupp, B.G., 2013. The way out: What we know and do not know about herpesvirus nuclear egress. Cellular Microbiology, 15(2), pp.170–178. Michael, K., Klupp, B.G., Mettenleiter, T.C. & Karger, A., 2006. Composition of pseudorabies virus particles lacking tegument protein US3, UL47, or UL49 or envelope glycoprotein E. Journal of virology, 80(3), pp.1332–1339. Mijatov, B., Cunningham, A.L. & Diefenbach, R.J., 2007. Residues F593 and E596 of HSV-1 tegument protein pUL36 (VP1/2) mediate binding of tegument protein pUL37. Virology, 368(1), pp.26–31. Milbradt, A.G., Kulkarni, M., Yi, T., Takeuchi, K., Sun, Z.-Y.J., Luna, R.E., Selenko, P., Näär, A.M. & Wagner, G., 2011. Structure of the VP16 transactivator target in the Mediator. Nature structural & molecular biology, 18(4), pp.410–415. Min, W.K., Cortes, U., Herceg, Z., Tong, W.M. & Wang, Z.Q., 2010. Deletion of the nuclear isoform of poly(ADP-ribose) glycohydrolase (PARG) reveals its function in DNA repair, genomic stability and tumorigenesis. Carcinogenesis, 31(12), pp.2058–2065. Miyazaki, D., Haruki, T., Takeda, S., Sasaki, S.I., Yakura, K., Terasaka, Y., Komatsu, N., Yamagami, S., Touge, H., Touge, C. & Inoue, Y., 2011. Herpes simplex virus type 1-induced transcriptional networks of corneal endothelial cells indicate antigen presentation function. Investigative Ophthalmology and Visual Science, 52, pp.4282–4293. Möhl, B.S., Böttcher, S., Granzow, H., Kuhn, J., Klupp, B.G. & Mettenleiter, T.C., 2009. Intracellular localization of the pseudorabies virus large tegument protein pUL36. Journal of virology, 83(19), pp.9641–9651. Mohni, K.N., Livingston, C.M., Cortez, D. & Weller, S.K., 2010. ATR and ATRIP are recruited to herpes simplex virus type 1 replication compartments even though ATR signaling is disabled. Journal of virology, 84(23), pp.12152–12164. Monette, A., Ajamian, L., López-Lastra, M. & Mouland, A.J., 2009. Human immunodeficiency virus type 1 (HIV-1) induces the cytoplasmic retention of heterogeneous nuclear ribonucleoprotein A1 by disrupting nuclear import. Implications for HIV-1 gene expression. Journal of Biological Chemistry, 284(45), pp.31350–31362. Monette, A., Panté, N. & Mouland, A.J., 2011. HIV-1 remodels the nuclear pore complex. Journal of Cell Biology, 193(4), pp.619–631. Moore, S.P., Rinckel, L.A. & Garfinkel, D.J., 1998. A Ty1 integrase nuclear localization signal required for retrotransposition. Molecular and cellular biology, 18(2), pp.1105–1114. Morgan, C., Rose, H.M. & Mednis, B., 1968. Electron microscopy of herpes simplex virus. I. Entry. Journal of virology, 2(5), pp.507–516. Morhenn, V.B., Lemperle, G. & Gallo, R.L., 2002. Phagocytosis of different particulate dermal filler substances by human macrophages and skin cells. Dermatologic Surgery, 28(6), pp.484–490. Morita, M., Kuba, K., Ichikawa, A., Nakayama, M., Katahira, J., Iwamoto, R., Watanebe, T., Sakabe, S., Daidoji, T., Nakamura, S., Kadowaki, A., Ohto, T., Nakanishi, H., Taguchi, R., Nakaya, T., Murakami, M., Yoneda, Y., Arai, H., Kawaoka, Y., Penninger, J.M., Arita, M. & Imai, Y., 2013. The lipid mediator protectin D1 inhibits influenza virus replication and improves severe influenza. Cell, 153(1), pp.112–25. Available at: http://www.sciencedirect.com/science/article/pii/S009286741300216X [Accessed March 11, 2015]. Morrison, E.E., Stevenson, A.J., Wang, Y.F. & Meredith, D.M., 1998. Differences in the intracellular localization and fate of herpes simplex virus tegument proteins early in the infection of Vero cells. Journal of General Virology, 79, pp.2517–2528. Mostafa, H.H., van Loben Sels, J.M. & Davido, D.J., 2015. HSV-1 upregulates p35, alters CDK-5 localization, and stimulates CDK-5 kinase activity during acute infection in neurons. Journal of Virology, (February), pp.JVI.00106–15. Available at: http://jvi.asm.org/lookup/doi/10.1128/JVI.00106-15. Mou, F., Forest, T. & Baines, J.D., 2007. US3 of herpes simplex virus type 1 encodes a promiscuous protein kinase that phosphorylates and alters localization of lamin A/C in infected cells. Journal of virology, 81(12), pp.6459–70. Available at: http://jvi.asm.org/content/81/12/6459.full [Accessed May 10, 2015]. Mullen, M. a, Ciufo, D.M. & Hayward, G.S., 1994. Mapping of intracellular localization domains and evidence for colocalization interactions between the IE110 and IE175 nuclear transactivator proteins of herpes simplex virus. Journal of virology, 68(5), pp.3250–3266. Munger, J. & Roizman, B., 2001. The US3 protein kinase of herpes simplex virus 1 mediates the posttranslational modification of BAD and prevents BAD-induced programmed cell death in the absence of other viral proteins. Proceedings of the National Academy of Sciences of the United States of America, 98(18), pp.10410–10415. Muranyi, W., Haas, J., Wagner, M., Krohne, G. & Koszinowski, U.H., 2002. Cytomegalovirus recruitment of cellular kinases to dissolve the nuclear lamina. Science (New York, N.Y.), 297(5582), pp.854–7. Available at: http://www.ncbi.nlm.nih.gov/pubmed/12161659 [Accessed May 10, 2015]. Muylaert, I. & Elias, P., 2007. Knockdown of DNA ligase IV/XRCC4 by RNA interference inhibits herpes simplex virus type I DNA replication. Journal of Biological Chemistry, 282(15), pp.10865–10872. Nagel, C.-H., Döhner, K., Fathollahy, M., Strive, T., Borst, E.M., Messerle, M. & Sodeik, B., 2008. Nuclear egress and envelopment of herpes simplex virus capsids analyzed with dual-color fluorescence HSV1(17+). Journal of virology, 82(6), pp.3109–3124. Nakamura, K., Kato, A., Kobayashi, J., Yanagihara, H., Sakamoto, S., Oliveira, D.V.N.P., Shimada, M., Tauchi, H., Suzuki, H., Tashiro, S., Zou, L. & Komatsu, K., 2011. Regulation of Homologous Recombination by RNF20-Dependent H2B Ubiquitination. Molecular Cell, 41(5), pp.515–528. Nakanishi, a., Shum, D., Morioka, H., Otsuka, E. & Kasamatsu, H., 2002. Interaction of the Vp3 Nuclear Localization Signal with the Importin 2/ Heterodimer Directs Nuclear Entry of Infecting Simian Virus 40. Journal of Virology, 76(18), pp.9368–9377. Available at: http://jvi.asm.org/cgi/doi/10.1128/JVI.76.18.9368-9377.2002. Nakanishi, A., Clever, J., Yamada, M., Li, P.P. & Kasamatsu, H., 1996. Association with capsid proteins promotes nuclear targeting of simian virus 40 DNA. Proceedings of the National Academy of Sciences of the United States of America, 93(1), pp.96–100. Nakanishi, A., Itoh, N., Li, P.P., Handa, H., Liddington, R.C. & Kasamatsu, H., 2007. Minor capsid proteins of simian virus 40 are dispensable

- 212 -

References

for nucleocapsid assembly and cell entry but are required for nuclear entry of the viral genome. Journal of virology, 81(8), pp.3778– 85. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1866110&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Naldini, L., Blömer, U., Gage, F.H., Trono, D. & Verma, I.M., 1996. Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proceedings of the National Academy of Sciences of the United States of America, 93(21), pp.11382–11388. Nellissery, J.K., Szczepaniak, R., Lamberti, C. & Weller, S.K., 2007. A putative leucine zipper within the herpes simplex virus type 1 UL6 protein is required for portal ring formation. Journal of virology, 81, pp.8868–8877. Nelson, L.M., Rose, R.C. & Moroianu, J., 2003. The L1 major capsid protein of human papillomavirus type 11 interacts with Kap ??2 and Kap ??3 nuclear import receptors. Virology, 306(1), pp.162–169. Nemeroff, M.E., Barabino, S.M., Li, Y., Keller, W. & Krug, R.M., 1998. Influenza virus NS1 protein interacts with the cellular 30 kDa subunit of CPSF and inhibits 3’end formation of cellular pre-mRNAs. Molecular cell, 1(7), pp.991–1000. Available at: http://www.ncbi.nlm.nih.gov/pubmed/9651582 [Accessed March 5, 2015]. Neumann, G., Castrucci, M.R. & Kawaoka, Y., 1997. Nuclear import and export of influenza virus nucleoprotein. Journal of virology, 71(12), pp.9690–9700. Neumann, G., Hughes, M.T. & Kawaoka, Y., 2000. Influenza A virus NS2 protein mediates vRNP nuclear export through NES-independent interaction with hCRM1. The EMBO journal, 19(24), pp.6751–8. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=305902&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Newcomb, W.W. & Brown, J.C., 1994. Induced extrusion of DNA from the capsid of herpes simplex virus type 1. Journal of virology, 68(1), pp.433–440. Newcomb, W.W. & Brown, J.C., 2010. Structure and capsid association of the herpesvirus large tegument protein UL36. Journal of virology, 84(18), pp.9408–9414. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2937621&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Newcomb, W.W. & Brown, J.C., 2009. Time-dependent transformation of the herpesvirus tegument. Journal of virology, 83(16), pp.8082– 8089. Newcomb, W.W., Homa, F.L. & Brown, J.C., 2006. Herpes simplex virus capsid structure: DNA packaging protein UL25 is located on the external surface of the capsid near the vertices. Journal of virology, 80(13), pp.6286–6294. Newcomb, W.W., Homa, F.L., Thomsen, D.R., Ye, Z. & Brown, J.C., 1994. Cell-free assembly of the herpes simplex virus capsid. Journal of virology, 68(9), pp.6059–6063. Newcomb, W.W., Jones, L.M., Dee, A., Chaudhry, F. & Brown, J.C., 2012. Role of a reducing environment in disassembly of the herpesvirus tegument. Virology, 431(1-2), pp.71–9. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22695308 [Accessed October 30, 2014]. Newcomb, W.W., Trus, B.L., Booy, F.P., Steven, A.C., Wall, J.S. & Brown, J.C., 1993. Structure of the herpes simplex virus capsid. Molecular composition of the pentons and the triplexes. Journal of molecular biology, 232(2), pp.499–511. Nguyen Ba, A.N., Pogoutse, A., Provart, N. & Moses, A.M., 2009. NLStradamus: a simple Hidden Markov Model for nuclear localization signal prediction. BMC bioinformatics, 10, p.202. Nicola, A., Hou, J., Major, E. & Straus, S., 2005. Herpes simplex virus type 1 enters human epidermal keratinocytes, but not neurons, via a pH- dependent endocytic pathway. Journal of virology, 79(12), pp.7609–7616. Available at: http://jvi.asm.org/content/79/12/7609.short. Nicola, A. V, Ponce de Leon, M., Xu, R., Hou, W., Whitbeck, J.C., Krummenacher, C., Montgomery, R.I., Spear, P.G., Eisenberg, R.J. & Cohen, G.H., 1998. Monoclonal antibodies to distinct sites on herpes simplex virus (HSV) glycoprotein D block HSV binding to HVEM. Journal of virology, 72(5), pp.3595–3601. Nicola, A. V & Straus, S.E., 2004. Cellular and viral requirements for rapid endocytic entry of herpes simplex virus. Journal of virology, 78(14), pp.7508–7517. De Nicola, F., Catena, V., Rinaldo, C., Bruno, T., Iezzi, S., Sorino, C., Desantis, a, Camerini, S., Crescenzi, M., Floridi, a, Passananti, C., Soddu, S. & Fanciulli, M., 2014. HIPK2 sustains apoptotic response by phosphorylating Che-1/AATF and promoting its degradation. Cell Death and Disease, 5(9), p.e1414. Available at: http://www.nature.com/doifinder/10.1038/cddis.2014.381. Nicoll, M.P., Proença, J.T. & Efstathiou, S., 2012. The molecular basis of herpes simplex virus latency. FEMS Microbiology Reviews, 36(3), pp.684–705. Nicolson, S.C. & Samulski, R.J., 2014. Recombinant adeno-associated virus utilizes host cell nuclear import machinery to enter the nucleus. Journal of virology, 88(8), pp.4132–44. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24478436. Nishimura, T., Kawai, N. & Ichihara, I., 1991. Interaction of endocytotic vacuoles with the inner nuclear membrane in simian virus 40 entry into CV-1 cell nucleus. Cell structure and function, 16(6), pp.441–445. de Noronha, C.M., Sherman, M.P., Lin, H.W., Cavrois, M. V, Moir, R.D., Goldman, R.D. & Greene, W.C., 2001. Dynamic disruptions in nuclear envelope architecture and integrity induced by HIV-1 Vpr. Science (New York, N.Y.), 294(5544), pp.1105–1108. O’Neill, R.E., Jaskunas, R., Blobel, G., Palese, P. & Moroianu, J., 1995. Nuclear import of influenza virus RNA can be mediated by viral nucleoprotein and transport factors required for protein import. Journal of Biological Chemistry, 270(39), pp.22701–22704. O’Neill, R.E., Talon, J. & Palese, P., 1998. The influenza virus NEP (NS2 protein) mediates the nuclear export of viral ribonucleoproteins. The EMBO journal, 17(1), pp.288–96. Available at:

- 213 -

References

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1170379&tool=pmcentrez&rendertype=abstract [Accessed April 30, 2015]. O’Reilly, A.J., Dacks, J.B. & Field, M.C., 2011. Evolution of the karyopherin-?? family of nucleocytoplasmic transport factors; ancient origins and continued specialization. PLoS ONE, 6(4). Ogawa-Goto, K., Tanaka, K., Gibson, W., Moriishi, E., Miura, Y., Kurata, T., Irie, S. & Sata, T., 2003. Microtubule network facilitates nuclear targeting of human cytomegalovirus capsid. Journal of virology, 77(15), pp.8541–8547. Ohkawa, T., Volkman, L.E. & Welch, M.D., 2010. Actin-based motility drives baculovirus transit to the nucleus and cell surface. Journal of Cell Biology, 190(2), pp.187–195. Ohta, A., Yamauchi, Y., Muto, Y., Kimura, H. & Nishiyama, Y., 2011. Herpes simplex virus type 1 UL14 tegument protein regulates intracellular compartmentalization of major tegument protein VP16. Virology journal, 8(1), p.365. Available at: http://www.virologyj.com/content/8/1/365. Ojala, P.M., Sodeik, B., Ebersold, M.W., Kutay, U. & Helenius, A., 2000. Herpes simplex virus type 1 entry into host cells: reconstitution of capsid binding and uncoating at the nuclear pore complex in vitro. Molecular and Cellular Biology, 20(13), pp.4922–4931. Ozawa, M., Fujii, K., Muramoto, Y., Yamada, S., Yamayoshi, S., Takada, A., Goto, H., Horimoto, T. & Kawaoka, Y., 2007. Contributions of two nuclear localization signals of influenza A virus nucleoprotein to viral replication. Journal of virology, 81(1), pp.30–41. Paine, P.L., Moore, L.C. & Horowitz, S.B., 1975. Nuclear envelope permeability. Nature, 254(5496), pp.109–114. Panté, N. & Kann, M., 2002. Nuclear pore complex is able to transport macromolecules with diameters of about 39 nm. Molecular biology of the cell, 13(2), pp.425–434. Pasdeloup, D., Beilstein, F., Roberts, a. P.E., McElwee, M., McNab, D. & Rixon, F.J., 2010. Inner tegument protein pUL37 of herpes simplex virus type 1 is involved in directing capsids to the trans-Golgi network for envelopment. Journal of General Virology, 91(9), pp.2145– 2151. Pasdeloup, D., Blondel, D., Isidro, A.L. & Rixon, F.J., 2009. Herpesvirus capsid association with the nuclear pore complex and viral DNA release involve the nucleoporin CAN/Nup214 and the capsid protein pUL25. Journal of virology, 83(13), pp.6610–6623. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2698519&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Pasdeloup, D., McElwee, M., Beilstein, F., Labetoulle, M. & Rixon, F.J., 2012. Herpesvirus Tegument Protein pUL37 Interacts with Dystonin/BPAG1 To Promote Capsid Transport on Microtubules during Egress. Journal of Virology, 87, pp.2857–2867. Available at: http://jvi.asm.org/cgi/doi/10.1128/JVI.02676-12. Patel, a, Hanson, J., McLean, T.I., Olgiate, J., Hilton, M., Miller, W.E. & Bachenheimer, S.L., 1998. Herpes simplex type 1 induction of persistent NF-kappa B nuclear translocation increases the efficiency of virus replication. Virology, 247(2), pp.212–222. Pellett, P. & Roizman, B., 2013. Herpesviridae, p 1802–1823. Fields virology, 6th ed. Lippincott Williams & Wilkins, Philadelphia, PA. Pemberton, L.F. & Paschal, B.M., 2005. Mechanisms of receptor-mediated nuclear import and nuclear export. Traffic (Copenhagen, Denmark), 6(3), pp.187–98. Available at: http://www.ncbi.nlm.nih.gov/pubmed/15702987 [Accessed October 15, 2014]. Lo Piano, A., Martínez-Jiménez, M.I., Zecchi, L. & Ayora, S., 2011. Recombination-dependent concatemeric viral DNA replication. Virus Research, 160(1-2), pp.1–14. Pinto, L.H. & Lamb, R. a., 2006. The M2 proton channels of influenza A and B viruses. Journal of Biological Chemistry, 281(14), pp.8997–9000. Placek, B.J. & Berger, S.L., 2010. Chromatin dynamics during herpes simplex virus-1 lytic infection. Biochimica et Biophysica Acta - Gene Regulatory Mechanisms, 1799(3-4), pp.223–227. Pleschka, S., Wolff, T., Ehrhardt, C., Hobom, G., Planz, O., Rapp, U.R. & Ludwig, S., 2001. Influenza virus propagation is impaired by inhibition of the Raf/MEK/ERK signalling cascade. Nature cell biology, 3(3), pp.301–5. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11231581 [Accessed May 10, 2015]. Pollard, V.W., Michael, W.M., Nakielny, S., Siomi, M.C., Wang, F. & Dreyfuss, G., 1996. A novel receptor-mediated nuclear protein import pathway. Cell, 86(6), pp.985–994. Popov, S., Rexach, M., Zybarth, G., Railing, N., Lee, M.A., Ratner, L., Lane, C.M., Moore, M.S., Blobel, G. & Bukrinsky, M., 1998. Viral protein R regulates nuclear import of the HIV-1 pre-integration complex. EMBO Journal, 17(4), pp.909–917. Porter, F.W. & Palmenberg, A.C., 2009. Leader-induced phosphorylation of nucleoporins correlates with nuclear trafficking inhibition by cardioviruses. Journal of virology, 83(4), pp.1941–51. Available at: http://jvi.asm.org/content/83/4/1941.abstract [Accessed March 22, 2015]. Porwal, M., Cohen, S., Snoussi, K., Popa-Wagner, R., Anderson, F., Dugot-Senant, N., Wodrich, H., Dinsart, C., Kleinschmidt, J.A., Panté, N. & Kann, M., 2013. Parvoviruses Cause Nuclear Envelope Breakdown by Activating Key Enzymes of Mitosis. PLoS Pathogens, 9(10). Post, L.E., Mackem, S. & Roizman, B., 1981. Regulation of alpha genes of herpes simplex virus: expression of chimeric genes produced by fusion of thymidine kinase with alpha gene promoters. Cell, 24(2), pp.555–565. Prchla, E., Plank, C., Wagner, E., Blaas, D. & Fuchs, R., 1995. Virus-mediated release of endosomal content in vitro: Different behavior of adenovirus and rhinovirus serotype 2. Journal of Cell Biology, 131(1), pp.111–123. Prokocimer, M., Margalit, A. & Gruenbaum, Y., 2006. The nuclear lamina and its proposed roles in tumorigenesis: Projection on the hematologic malignancies and future targeted therapy. Journal of Structural Biology, 155(2), pp.351–360. Pulliam, K.F., Fasken, M.B., McLane, L.M., Pulliam, J. V & Corbett, A.H., 2009. The classical nuclear localization signal receptor, importin-alpha, is required for efficient transition through the G1/S stage of the cell cycle in Saccharomyces cerevisiae. Genetics, 181(1), pp.105–18. Available at: http://www.genetics.org/content/181/1/105.full [Accessed May 10, 2015].

- 214 -

References

Purves, F.C., Spector, D. & Roizman, B., 1991. The herpes simplex virus 1 protein kinase encoded by the US3 gene mediates posttranslational modification of the phosphoprotein encoded by the UL34 gene. Journal of virology, 65(11), pp.5757–5764. Radivojac, P., Vacic, V., Haynes, C., Cocklin, R.R., Mohan, A., Heyen, J.W., Goebl, M.G. & Iakoucheva, L.M., 2010. Identification, analysis, and prediction of protein ubiquitination sites. Proteins, 78(2), pp.365–380. Radtke, K., Kieneke, D., Wolfstein, A., Michael, K., Steffen, W., Scholz, T., Karger, A. & Sodeik, B., 2010. Plus- and minus-end directed microtubule motors bind simultaneously to herpes simplex virus capsids using different inner tegument structures. PLoS Pathogens, 6(7), pp.1–20. Rascle, A., Neumann, T., Raschta, A.S., Neumann, A., Heining, E., Kastner, J. & Witzgall, R., 2009. The LIM-homeodomain transcription factor LMX1B regulates expression of NF-kappa B target genes. Experimental Cell Research, 315(1), pp.76–96. Reichelt, R., Holzenburg, a, Buhle, E.L., Jarnik, M., Engel, a & Aebi, U., 1990. Correlation between Structure and Mass-Distribution of the Nuclear-Pore Complex and of Distinct Pore Complex Components. Journal of Cell Biology, 110(4), pp.883–894. Available at: ://A1990CX59200002. Reid, S.P., Leung, L.W., Hartman, A.L., Martinez, O., Shaw, M.L., Carbonnelle, C., Volchkov, V.E., Nichol, S.T. & Basler, C.F., 2006. Ebola virus VP24 binds karyopherin alpha1 and blocks STAT1 nuclear accumulation. Journal of virology, 80(11), pp.5156–67. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1472181&tool=pmcentrez&rendertype=abstract [Accessed March 2, 2015]. Reid, S.P., Valmas, C., Martinez, O., Sanchez, F.M. & Basler, C.F., 2007. Ebola virus VP24 proteins inhibit the interaction of NPI-1 subfamily karyopherin alpha proteins with activated STAT1. Journal of virology, 81(24), pp.13469–77. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2168840&tool=pmcentrez&rendertype=abstract [Accessed April 5, 2015]. Reynolds, A.E., Wills, E.G., Roller, R.J., Ryckman, B.J. & Baines, J.D., 2002. Ultrastructural Localization of the Herpes Simplex Virus Type 1 U L 31 , U L 34 , and U S 3 Proteins Suggests Specific Roles in Primary Envelopment and Egress of Nucleocapsids. Journal of Virology, 76(17), pp.8939–8952. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=136992&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Ricour, C., Borghese, F., Sorgeloos, F., Hato, S. V, van Kuppeveld, F.J.M. & Michiels, T., 2009. Random mutagenesis defines a domain of Theiler’s virus leader protein that is essential for antagonism of nucleocytoplasmic trafficking and cytokine gene expression. Journal of virology, 83(21), pp.11223–32. Available at: http://jvi.asm.org/content/83/21/11223.abstract [Accessed April 12, 2015]. Riddick, G. & Macara, I.G., 2007. The adapter importin-alpha provides flexible control of nuclear import at the expense of efficiency. Molecular systems biology, 3(118), p.118. Rihs, H.P., Jans, D.A., Fan, H. & Peters, R., 1991. The rate of nuclear cytoplasmic protein transport is determined by the casein kinase II site flanking the nuclear localization sequence of the SV40 T-antigen. The EMBO journal, 10(3), pp.633–639. Rixon, F.J., Addison, C., McGregor, A., Macnab, S.J., Nicholson, P., Preston, V.G. & Tatman, J.D., 1996. Multiple interactions control the intracellular localization of the herpes simplex virus type 1 capsid proteins. The Journal of general virology, 77 ( Pt 9), pp.2251–2260. Roberts, A.P.E., Abaitua, F., O’Hare, P., McNab, D., Rixon, F.J. & Pasdeloup, D., 2009. Differing roles of inner tegument proteins pUL36 and pUL37 during entry of herpes simplex virus type 1. Journal of virology, 83(1), pp.105–116. Rode, K., Döhner, K., Binz, A., Glass, M., Strive, T., Bauerfeind, R. & Sodeik, B., 2011. Uncoupling uncoating of herpes simplex virus genomes from their nuclear import and gene expression. Journal of virology, 85(9), pp.4271–83. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3126263&tool=pmcentrez&rendertype=abstract [Accessed October 21, 2014]. Roe, T., Reynolds, T.C., Yu, G. & Brown, P.O., 1993. Integration of murine leukemia virus DNA depends on mitosis. The EMBO journal, 12(5), pp.2099–2108. Roizman, B., 2011. The checkpoints of viral gene expression in productive and latent infection: the role of the HDAC/CoREST/LSD1/REST repressor complex. Journal of virology, 85(15), pp.7474–7482. Roizman, B., Knipe, D.M. & Whitley, R.J., 2013. Herpes simplex viruses, p 1823–1897. Fields virology, 6th ed. Lippincott Williams & Wilkins, Philadelphia, PA. Roizman, B. & Whitley, R.J., 2013. An inquiry into the molecular basis of HSV latency and reactivation. Annual review of microbiology, 67, pp.355–74. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24024635. Roller, R.J. & Fetters, R., 2015. The Herpes Simplex Virus 1 UL51 Protein Interacts with the UL7 Protein and Plays a Role in Its Recruitment into the Virion. Journal of Virology, 89(6), pp.3112–3122. Available at: http://jvi.asm.org/lookup/doi/10.1128/JVI.02799-14. Roman, N., Christie, M., Swarbrick, C.M.D., Kobe, B. & Forwood, J.K., 2013. Structural Characterisation of the Nuclear Import Receptor Importin Alpha in Complex with the Bipartite NLS of Prp20. PLoS ONE, 8(12), p.e82038. Available at: http://dx.plos.org/10.1371/journal.pone.0082038. Rossman, J.S., Leser, G.P. & Lamb, R. a., 2012. Filamentous Influenza Virus Enters Cells via Macropinocytosis. Journal of Virology, 86(20), pp.10950–10960. Available at: http://dx.doi.org/10.1128/JVI.05992-11. Le Rouzic, E., Mousnier, A., Rustum, C., Stutz, F., Hallberg, E., Dargemont, C. & Benichou, S., 2002. Docking of HIV-1 vpr to the nuclear envelope is mediated by the interaction with the nucleoporin hCG1. Journal of Biological Chemistry, 277(47), pp.45091–45098. Ruben, G.J., Kirkland, J.G., MacDonough, T., Chen, M., Dubey, R.N., Gartenberg, M.R. & Kamakaka, R.T., 2011. Nucleoporin mediated nuclear positioning and silencing of HMR. PLoS ONE, 6(7).

- 215 -

References

Saeed, M.F., Kolokoltsov, A. a., Albrecht, T. & Davey, R. a., 2010. Cellular entry of ebola virus involves uptake by a macropinocytosis-like mechanism and subsequent trafficking through early and late endosomes. PLoS Pathogens, 6(9). Salsman, J., Zimmerman, N., Chen, T., Domagala, M. & Frappier, L., 2008. Genome-wide screen of three herpesviruses for protein subcellular localization and alteration of PML nuclear bodies. PLoS Pathogens, 4(7). Saphire, a C., Guan, T., Schirmer, E.C., Nemerow, G.R. & Gerace, L., 2000. Nuclear import of adenovirus DNA in vitro involve the nuclear protin import pathway and Hsc70. Journal of Biological Chemistry, 275(6), pp.4298–4304. Satterly, N., Tsai, P.-L., van Deursen, J., Nussenzveig, D.R., Wang, Y., Faria, P.A., Levay, A., Levy, D.E. & Fontoura, B.M.A., 2007. Influenza virus targets the mRNA export machinery and the nuclear pore complex. Proceedings of the National Academy of Sciences of the United States of America, 104(6), pp.1853–8. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1794296&tool=pmcentrez&rendertype=abstract [Accessed March 30, 2015]. Schaffer, P.A., Carter, V.C. & Timbury, M.C., 1978. Collaborative complementation study of temperature-sensitive mutants of herpes simplex virus types 1 and 2. J Virol, 27(3), pp.490–504. Schaller, T., Ocwieja, K.E., Rasaiyaah, J., Price, A.J., Brady, T.L., Roth, S.L., Hué, S., Fletcher, A.J., Lee, K., KewalRamani, V.N., Noursadeghi, M., Jenner, R.G., James, L.C., Bushman, F.D. & Towers, G.J., 2011. HIV-1 capsid-cyclophilin interactions determine nuclear import pathway, integration targeting and replication efficiency. PLoS Pathogens, 7(12). Schang, L.M., Phillips, J. & Schaffer, P. a, 1998. Requirement for cellular cyclin-dependent kinases in herpes simplex virus replication and transcription. Journal of virology, 72(7), pp.5626–5637. Schang, L.M., Rosenberg, A. & Schaffer, P.A., 1999. Transcription of herpes simplex virus immediate-early and early genes is inhibited by roscovitine, an inhibitor specific for cellular cyclin-dependent kinases. Journal of virology, 73(3), pp.2161–2172. Schipke, J., Pohlmann, a., Diestel, R., Binz, a., Rudolph, K., Nagel, C.-H., Bauerfeind, R. & Sodeik, B., 2012. The C Terminus of the Large Tegument Protein pUL36 Contains Multiple Capsid Binding Sites That Function Differently during Assembly and Cell Entry of Herpes Simplex Virus. Journal of Virology, 86(7), pp.3682–3700. Schlieker, C., Korbel, G. a, Kattenhorn, L.M. & Ploegh, H.L., 2005. A deubiquitinating activity is conserved in the large tegument protein of the herpesviridae. Journal of virology, 79(24), pp.15582–15585. Schmitz, A., Schwarz, A., Foss, M., Zhou, L., Rabe, B., Hoellenriegel, J., Stoeber, M., Panté, N. & Kann, M., 2010. Nucleoporin 153 arrests the nuclear import of hepatitis B virus capsids in the nuclear basket. PLoS pathogens, 6(1), p.e1000741. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2813275&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Schoenfeld, A.R., Apgar, S., Dolios, G., Wang, R. & Aaronson, S. a, 2004. BRCA2 is ubiquitinated in vivo and interacts with USP11, a deubiquitinating enzyme that exhibits prosurvival function in the cellular response to DNA damage. Molecular and cellular biology, 24(17), pp.7444–7455. Schoggins, J.W., Wilson, S.J., Panis, M., Murphy, M.Y., Jones, C.T., Bieniasz, P. & Rice, C.M., 2011. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature, 472(7344), pp.481–485. Available at: http://dx.doi.org/10.1038/nature09907. Schulz, K.S., Klupp, B.G., Granzow, H., Paßvogel, L. & Mettenleiter, T.C., 2015. Herpesvirus nuclear egress: Pseudorabies Virus can simultaneously induce nuclear envelope breakdown and exit the nucleus via the envelopment–deenvelopment-pathway. Virus Research, pp.1–11. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0168170215000544. Schulz, K.S., Liu, X., Klupp, B.G., Granzow, H., Cohen, J.I. & Mettenleiter, T.C., 2014. Pseudorabies virus pUL46 induces activation of ERK1/2 and regulates herpesvirus-induced nuclear envelope breakdown. Journal of virology, 88, pp.6003–11. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24623429. Schumacher, A.J., Mohni, K.N., Kan, Y., Hendrickson, E. a., Stark, J.M. & Weller, S.K., 2012. The HSV-1 Exonuclease, UL12, Stimulates Recombination by a Single Strand Annealing Mechanism. PLoS Pathogens, 8(8). Sciortino, M.T., Suzuki, M., Taddeo, B. & Roizman, B., 2001. RNAs extracted from herpes simplex virus 1 virions: apparent selectivity of viral but not cellular RNAs packaged in virions. Journal of virology, 75(17), pp.8105–8116. Scott, E. & O’Hare, P., 2001. Fate of the inner nuclear membrane protein lamin B receptor and nuclear lamins in herpes simplex virus type 1 infection. Journal of virology, 75(18), pp.8818–8830. Available at: http://jvi.asm.org/content/75/18/8818.short. Scott, M.S., Boisvert, F.M., McDowall, M.D., Lamond, A.I. & Barton, G.J., 2010. Characterization and prediction of protein nucleolar localization sequences. Nucleic Acids Research, 38(21), pp.7388–7399. Scrima, N., Lepault, J., Boulard, Y., Pasdeloup, D., Bressanelli, S. & Roche, S., 2015. Insights into Herpesvirus Tegument Organization from Structural Analyses of HSV-1 UL36 Central 970 Residues. Journal of Biological Chemistry, 1, p.jbc.M114.612838. Available at: http://www.jbc.org/lookup/doi/10.1074/jbc.M114.612838. Seth, P., 1994. Adenovirus-dependent release of choline from plasma membrane vesicles at an acidic pH is mediated by the penton base protein. Journal of virology, 68(2), pp.1204–1206. Seth, P., Fitzgerald, D.J., Willingham, M.C. & Pastan, I., 1984. Role of a low-pH environment in adenovirus enhancement of the toxicity of a Pseudomonas exotoxin-epidermal growth factor conjugate. Journal of virology, 51(3), pp.650–655. Shabman, R.S., Gulcicek, E.E., Stone, K.L. & Basler, C.F., 2011. The Ebola virus VP24 protein prevents hnRNP C1/C2 binding to karyopherin α1 and partially alters its nuclear import. The Journal of infectious diseases, 204 Suppl , pp.S904–10. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3189985&tool=pmcentrez&rendertype=abstract [Accessed January 21,

- 216 -

References

2015]. Shapira, S.D., Gat-Viks, I., Shum, B.O. V, Dricot, A., de Grace, M.M., Wu, L., Gupta, P.B., Hao, T., Silver, S.J., Root, D.E., Hill, D.E., Regev, A. & Hacohen, N., 2009. A Physical and Regulatory Map of Host-Influenza Interactions Reveals Pathways in H1N1 Infection. Cell, 139(7), pp.1255–1267. Shimizu, T., Takizawa, N., Watanabe, K., Nagata, K. & Kobayashi, N., 2011. Crucial role of the influenza virus NS2 (NEP) C-terminal domain in M1 binding and nuclear export of vRNP. FEBS letters, 585(1), pp.41–6. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21081124 [Accessed April 19, 2015]. Shirata, N., Kudoh, A., Daikoku, T., Tatsumi, Y., Fujita, M., Kiyono, T., Sugaya, Y., Isomura, H., Ishizaki, K. & Tsurumi, T., 2005. Activation of ataxia telangiectasia-mutated DNA damage checkpoint signal transduction elicited by herpes simplex virus infection. Journal of Biological Chemistry, 280(34), pp.30336–30341. Shu, M., Taddeo, B. & Roizman, B., 2013. The nuclear-cytoplasmic shuttling of virion host shutoff RNase is enabled by pUL47 and an embedded nuclear export signal and defines the sites of degradation of AU-rich and stable cellular mRNAs. Journal of virology, 87(24), pp.13569–78. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24109211. Simpson-holley, M., Baines, J., Roller, R. & Knipe, D.M., 2004. Herpes Simplex Virus 1 U L 31 and U L 34 Gene Products Promote the Late Maturation of Viral Replication Compartments to the Nuclear Periphery. Society, 78(11), pp.5591–5600. Singh, S.K. & Ruzek, D., 2013. Neuroviral Infections: General Principles and DNA Viruses, CRC Press/Taylor & Francis. Available at: https://books.google.de/books?id=5_zn69N8nZUC. Siomi, H. & Dreyfuss, G., 1995. A nuclear localization domain in the hnRNP A1 protein. Journal of Cell Biology, 129(3), pp.551–560. Skehel, J.J. & Wiley, D.C., 2000. M Echanisms of V Iral M Embrane F Usion and I Ts I Nhibition. Annual review of biochemistry, 69, pp.531– 569. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11395423. Smith, J.G., Cassany, A., Gerace, L., Ralston, R. & Nemerow, G.R., 2008. Neutralizing antibody blocks adenovirus infection by arresting microtubule-dependent cytoplasmic transport. Journal of virology, 82(13), pp.6492–6500. Smith, M.C., Boutell, C. & Davido, D.J., 2011. HSV-1 ICP0: paving the way for viral replication. , 6(4), pp.421–429. Available at: http://dx.doi.org/10.2217/FVL.11.24. Smith, S., Reuven, N., Mohni, K.N., Schumacher, A.J. & Weller, S.K., 2014. Structure of the herpes simplex virus 1 genome: manipulation of nicks and gaps can abrogate infectivity and alter the cellular DNA damage response. Journal of virology, 88(17), pp.10146–56. Available at: http://jvi.asm.org/content/early/2014/06/24/JVI.01723-14.abstract. Snow, C.M., Senior, A. & Gerace, L., 1987. Monoclonal antibodies identify a group of nuclear pore complex glycoproteins. Journal of Cell Biology, 104(5), pp.1143–1156. Sodeik, B., Ebersold, M.W. & Helenius, A., 1997. Microtubule-mediated Transport of Incoming Herpes Simplex Virus 1 Capsids to the Nucleus. The Journal of Cell Biology, 136(5), pp.1007–1021. Sorokin, a. V., Kim, E.R. & Ovchinnikov, L.P., 2007. Nucleocytoplasmic transport of proteins. Biochemistry (Moscow), 72(13), pp.1439–1457. Available at: http://link.springer.com/10.1134/S0006297907130032. Spear, P.G. & Roizman, B., 1972. Proteins specified by herpes simplex virus. V. Purification and structural proteins of the herpesvirion. Journal of virology, 9(I), pp.143–159. Speese, S.D., Ashley, J., Jokhi, V., Nunnari, J., Barria, R., Li, Y., Ataman, B., Koon, A., Chang, Y.T., Li, Q., Moore, M.J. & Budnik, V., 2012. Nuclear envelope budding enables large ribonucleoprotein particle export during synaptic Wnt signaling. Cell, 149(4), pp.832–846. Available at: http://dx.doi.org/10.1016/j.cell.2012.03.032. Stoffler, D., Fahrenkrog, B. & Aebi, U., 1999. The nuclear pore complex: from molecular architecture to functional dynamics. Current opinion in cell biology, 11(3), pp.391–401. Strunze, S., Engelke, M.F., Wang, I.-H., Puntener, D., Boucke, K., Schleich, S., Way, M., Schoenenberger, P., Burckhardt, C.J. & Greber, U.F., 2011. Kinesin-1-mediated capsid disassembly and disruption of the nuclear pore complex promote virus infection. Cell host & microbe, 10(3), pp.210–23. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21925109 [Accessed October 30, 2014]. Strunze, S., Trotman, L.C., Boucke, K. & Greber, U.F., 2005. Nuclear Targeting of Adenovirus Type 2 Requires CRM1-mediated Nuclear Export □. , 16(June), pp.2999–3009. Subramanian, R.P. & Geraghty, R.J., 2007. Herpes simplex virus type 1 mediates fusion through a hemifusion intermediate by sequential activity of glycoproteins D, H, L, and B. Proceedings of the National Academy of Sciences of the United States of America, 104(8), pp.2903–2908. Suzuki, T., Tsuzuku, J., Hayashi, A., Shiomi, Y., Iwanari, H., Mochizuki, Y., Hamakubo, T., Kodama, T., Nishitani, H., Masai, H. & Yamamoto, T., 2012. Inhibition of DNA damage-induced apoptosis through Cdc7-mediated stabilization of Tob. Journal of Biological Chemistry, 287(48), pp.40256–40265. Svobodova, S., Bell, S. & Crump, C.M., 2012. Analysis of the Interaction between the Essential Herpes Simplex Virus 1 Tegument Proteins VP16 and VP1/2. Journal of Virology, 86(1), pp.473–483. Szilágyi, J.F. & Cunningham, C., 1991. Identification and characterization of a novel non-infectious herpes simplex virus-related particle. The Journal of general virology, 72 ( Pt 3), pp.661–8. Available at: http://europepmc.org/abstract/med/1848601 [Accessed May 10, 2015]. Takakuwa, H., Goshima, F., Koshizuka, T., Murata, T., Daikoku, T. & Nishiyama, Y., 2001. Herpes simplex virus encodes a virion-associated protein which promotes long cellular processes in over-expressing cells. Genes to Cells, 6(11), pp.955–966. Takeyama, K., Aguiar, R.C.T., Gu, L., He, C., Freeman, G.J., Kutok, J.L., Aster, J.C. & Shipp, M.A., 2003. The BAL-binding protein BBAP and related deltex family members exhibit ubiquitin-protein isopeptide ligase activity. Journal of Biological Chemistry, 278(24), pp.21930–

- 217 -

References

21937. Tandon, R., Mocarski, E. & Conway, J., 2015. The A, B, Cs of Herpesvirus Capsids. Viruses, 7, pp.899–914. Available at: http://www.mdpi.com/1999-4915/7/3/899/. Taylor, T.J. & Knipe, D.M., 2004. Proteomics of herpes simplex virus replication compartments: association of cellular DNA replication, repair, recombination, and chromatin remodeling proteins with ICP8. Journal of virology, 78(11), pp.5856–5866. Thierry-Mieg, D. & Thierry-Mieg, J., 2006. AceView: a comprehensive cDNA-supported gene and transcripts annotation. Genome biology, 7 Suppl 1(Suppl 1), pp.S12.1–14. Thomas, J.A., Ott, D.E. & Gorelick, R.J., 2007. Efficiency of human immunodeficiency virus type 1 postentry infection processes: evidence against disproportionate numbers of defective virions. Journal of virology, 81(8), pp.4367–4370. Thompson, R.L., Williams, R.W., Kotb, M. & Sawtell, N.M., 2014. A forward phenotypically driven unbiased genetic analysis of host genes that moderate herpes simplex virus virulence and stromal keratitis in mice. PloS one, 9(3), p.e92342. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24651695. Thomsen, D.R., Roof, L.L. & Homa, F.L., 1994. Assembly of herpes simplex virus (HSV) intermediate capsids in insect cells infected with recombinant baculoviruses expressing HSV capsid proteins. Journal of virology, 68(4), pp.2442–2457. Thomson, T.M. & Guerra-Rebollo, M., 2010. Ubiquitin and SUMO signalling in DNA repair. Biochemical Society transactions, 38(Pt 1), pp.116– 131. Trotman, L.C., Mosberger, N., Fornerod, M., Stidwill, R.P. & Greber, U.F., 2001. Import of adenovirus DNA involves the nuclear pore complex receptor CAN/Nup214 and histone H1. Nature cell biology, 3(December), pp.1092–1100. Trus, B.L., Cheng, N., Newcomb, W.W., Homa, L., Brown, J.C. & Steven, A.C., 2004. Structure and Polymorphism of the UL6 Portal Protein of Herpes Simplex Virus Type 1 Structure and Polymorphism of the UL6 Portal Protein of Herpes Simplex Virus Type 1. Journal of virology, 78(22), pp.12668–12671. Unterholzner, L., Keating, S.E., Baran, M., Horan, K.A., Jensen, S.B., Sharma, S., Sirois, C.M., Jin, T., Latz, E., Xiao, T.S., Fitzgerald, K.A., Paludan, S.R. & Bowie, A.G., 2010. IFI16 is an innate immune sensor for intracellular DNA. Nature immunology, 11(11), pp.997–1004. Available at: http://dx.doi.org/10.1038/ni.1932. Urbé, S., 2005. Ubiquitin and endocytic protein sorting. Essays in biochemistry, 41, pp.81–98. Valle-Casuso, J.C., Di Nunzio, F., Yang, Y., Reszka, N., Lienlaf, M., Arhel, N., Perez, P., Brass, A.L. & Diaz-Griffero, F., 2012. TNPO3 Is Required for HIV-1 Replication after Nuclear Import but prior to Integration and Binds the HIV-1 Core. Journal of Virology, 86(10), pp.5931– 5936. Vasicova, P., Stradalova, V., Halada, P., Hasek, J. & Malcova, I., 2013. Nuclear Import of Chromatin Remodeler Isw1 Is Mediated by Atypical Bipartite cNLS and Classical Import Pathway. Traffic, 14(2), pp.176–193. Ventii, K.H. & Wilkinson, K.D., 2008. Protein partners of deubiquitinating enzymes. The Biochemical journal, 414(2), pp.161–175. Vittone, V., Diefenbach, E., Triffett, D., Mark, W., Cunningham, A.L., Russell, J., Douglas, M.W. & Diefenbach, R.J., 2005. Determination of Interactions between Tegument Proteins of Herpes Simplex Virus Type 1 Determination of Interactions between Tegument Proteins of Herpes Simplex Virus Type 1. , 79(15), pp.9566–9571. Vodicka, M.A., Koepp, D.M., Silver, P.A. & Emerman, M., 1998. HIV-1 Vpr interacts with the nuclear transport pathway to promote macrophage infection. Genes and Development, 12(2), pp.175–185. Van De Vosse, D.W., Wan, Y., Lapetina, D.L., Chen, W.M., Chiang, J.H., Aitchison, J.D. & Wozniak, R.W., 2013. A role for the nucleoporin Nup170p in chromatin structure and gene silencing. Cell, 152(5), pp.969–983. Available at: http://dx.doi.org/10.1016/j.cell.2013.01.049. Wang, N., Baldi, P.F. & Gaut, B.S., 2007. Phylogenetic analysis, genome evolution and the rate of gene gain in the Herpesviridae. Molecular Phylogenetics and Evolution, 43(3), pp.1066–1075. Wang, P., Palese, P. & O’Neill, R.E., 1997. The NPI-1/NPI-3 (karyopherin alpha) binding site on the influenza a virus nucleoprotein NP is a nonconventional nuclear localization signal. Journal of virology, 71(3), pp.1850–1856. Wang, R. & Brattain, M.G., 2007. The maximal size of protein to diffuse through the nuclear pore is larger than 60 kDa. FEBS Letters, 581(17), pp.3164–3170. Wang, S., Wang, K., Li, J. & Zheng, C., 2013. Herpes simplex virus 1 ubiquitin-specific protease UL36 inhibits beta interferon production by deubiquitinating TRAF3. Journal of virology, 87(21), pp.11851–60. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3807349&tool=pmcentrez&rendertype=abstract. Watanabe, D., Ushijima, Y., Goshima, F., Takakuwa, H., Tomita, Y. & Nishiyama, Y., 2000. Identification of nuclear export signal in UL37 protein of herpes simplex virus type 2. Biochemical and biophysical research communications, 276, pp.1248–1254. Watanabe, K., Takizawa, N., Katoh, M., Hoshida, K., Kobayashi, N. & Nagata, K., 2001. Inhibition of nuclear export of ribonucleoprotein complexes of influenza virus by leptomycin B. Virus research, 77(1), pp.31–42. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11451485 [Accessed May 10, 2015]. van der Watt, P.J., Ngarande, E. & Leaner, V.D., 2011. Overexpression of Kpn??1 and Kpn??2 importin proteins in cancer derives from deregulated E2F activity. PLoS ONE, 6(11). Weber, F., Kochs, G., Gruber, S. & Haller, O., 1998. A classical bipartite nuclear localization signal on Thogoto and influenza A virus nucleoproteins. Virology, 250(1), pp.9–18. Weinheimer, S.P., Boyd, B. a, Durham, S.K., Resnick, J.L. & O’Boyle, D.R., 1992. Deletion of the VP16 open reading frame of herpes simplex virus type 1. Journal of virology, 66(1), pp.258–269.

- 218 -

References

Wild, P., Engels, M., Senn, C., Tobler, K., Ziegler, U., Schraner, E.M., Loepfe, E., Ackermann, M., Mueller, M. & Walther, P., 2005. Impairment of nuclear pores in bovine herpesvirus 1-infected MDBK cells. Journal of virology, 79(2), pp.1071–83. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=538577&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Wild, P., Senn, C., Manera, C.L., Sutter, E., Schraner, E.M., Tobler, K., Ackermann, M., Ziegler, U., Lucas, M.S. & Kaech, A., 2009. Exploring the nuclear envelope of herpes simplex virus 1-infected cells by high-resolution microscopy. Journal of virology, 83(1), pp.408–19. Available at: http://jvi.asm.org/content/83/1/408.full [Accessed May 10, 2015]. Wilkinson, D.E. & Weller, S.K., 2006. Herpes simplex virus type I disrupts the ATR-dependent DNA-damage response during lytic infection. Journal of cell science, 119(Pt 13), pp.2695–2703. Wilkinson, D.E. & Weller, S.K., 2004. Recruitment of Cellular Recombination and Repair Proteins to Sites of Herpes Simplex Virus Type 1 DNA Replication Is Dependent on the Composition of Viral Proteins within Prereplicative Sites and Correlates with the Induction of the DNA Damage Response. Journal of Virology, 78(9), pp.4783–4796. Available at: http://jvi.asm.org/content/78/9/4783.long [Accessed April 9, 2015]. Wilkinson, D.E. & Weller, S.K., 2003. The role of DNA recombination in herpes simplex virus DNA replication. IUBMB life, 55(8), pp.451–458. Wilson, K.L. & Berk, J.M., 2010. The nuclear envelope at a glance. Journal of cell science, 123(Pt 12), pp.1973–1978. Wisner, T.W., Wright, C.C., Kato, A., Kawaguchi, Y., Mou, F., Baines, J.D., Roller, R.J. & Johnson, D.C., 2009. Herpesvirus gB-induced fusion between the virion envelope and outer nuclear membrane during virus egress is regulated by the viral US3 kinase. Journal of virology, 83(7), pp.3115–26. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2655551&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Wodrich, H., Cassany, A., D’Angelo, M. a, Guan, T., Nemerow, G. & Gerace, L., 2006. Adenovirus core protein pVII is translocated into the nucleus by multiple import receptor pathways. Journal of virology, 80(19), pp.9608–9618. Wolfstein, A., Nagel, C.H., Radtke, K., Döhner, K., Allan, V.J. & Sodeik, B., 2006. The inner tegument promotes herpes simplex virus capsid motility along microtubules in vitro. Traffic, 7(6), pp.227–237. Wong, Y.H., Lee, T.Y., Liang, H.K., Huang, C.M., Wang, T.Y., Yang, Y.H., Chu, C.H., Huang, H. Da, Ko, M.T. & Hwang, J.K., 2007. KinasePhos 2.0: A web server for identifying protein kinase-specific phosphorylation sites based on sequences and coupling patterns. Nucleic Acids Research, 35(SUPPL.2). Wright, C.C., Wisner, T.W., Hannah, B.P., Eisenberg, R.J., Cohen, G.H. & Johnson, D.C., 2009. Fusion between perinuclear virions and the outer nuclear membrane requires the fusogenic activity of herpes simplex virus gB. Journal of virology, 83(22), pp.11847–56. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2772685&tool=pmcentrez&rendertype=abstract [Accessed May 10, 2015]. Wu, W.W.H., Sun, Y.-H.B. & Panté, N., 2007. Nuclear import of influenza A viral ribonucleoprotein complexes is mediated by two nuclear localization sequences on viral nucleoprotein. Virology journal, 4, p.49. Xiao, C.Y., Hübner, S. & Jans, D.A., 1997. SV40 large tumor antigen nuclear import is regulated by the double- stranded DNA-dependent protein kinase site (Serine 120) flanking the nuclear localization sequence. Journal of Biological Chemistry, 272(35), pp.22191–22198. Xing, J., Wang, S., Li, Y., Guo, H., Zhao, L., Pan, W., Lin, F., Zhu, H., Wang, L.L., Li, M., Wang, L.L. & Zheng, C., 2011. Characterization of the subcellular localization of herpes simplex virus type 1 proteins in living cells. Medical Microbiology and Immunology, 200, pp.61–68. Xu, F., Schillinger, J.A., Sternberg, M.R., Johnson, R.E., Lee, F.K., Nahmias, A.J. & Markowitz, L.E., 2002. Seroprevalence and coinfection with herpes simplex virus type 1 and type 2 in the United States, 1988-1994. The Journal of infectious diseases, 185(8), pp.1019–1024. Xu, F., Sternberg, M.R., Kottiri, B.J., McQuillan, G.M., Lee, F.K., Nahmias, A.J., Berman, S.M. & Markowitz, L.E., 2006. Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States. JAMA : the journal of the American Medical Association, 296(8), pp.964– 973. Xu, H., Yao, L., Lu, S. & Qi, Y., 2007. Host filamentous actin is associated with Heliothis armigera single nucleopolyhedrosis virus (HaSNPV) nucleocapsid transport to the host nucleus. Current Microbiology, 54(3), pp.199–206. Xue, Y., Liu, Z., Cao, J., Ma, Q., Gao, X., Wang, Q., Jin, C., Zhou, Y., Wen, L. & Ren, J., 2011. GPS 2.1: Enhanced prediction of kinase-specific phosphorylation sites with an algorithm of motif length selection. Protein Engineering, Design and Selection, 24(3), pp.255–260. Xue, Y., Ren, J., Gao, X., Jin, C., Wen, L. & Yao, X., 2008. GPS 2.0, a tool to predict kinase-specific phosphorylation sites in hierarchy. Molecular & cellular proteomics : MCP, 7(9), pp.1598–1608. Yamada, H., Jiang, Y.M., Oshima, S., Daikoku, T., Yamashita, Y., Tsurumi, T. & Nishiyama, Y., 1998. Characterization of the UL55 gene product of herpes simplex virus type 2. The Journal of general virology, 79 ( Pt 8), pp.1989–1995. Yamada, M. & Kasamatsu, H., 1993. Role of nuclear pore complex in simian virus 40 nuclear targeting. Journal of virology, 67(1), pp.119–130. Yamaguchi, T., Kimura, J., Miki, Y. & Yoshida, K., 2007. The deubiquitinating enzyme USP11 controls an I??B kinase ?? (IKK??)-p53 signaling pathway in response to Tumor Necrosis Factor ?? (TNF??). Journal of Biological Chemistry, 282(47), pp.33943–33948. Yamashita, M. & Emerman, M., 2004. Capsid is a dominant determinant of retrovirus infectivity in nondividing cells. Journal of virology, 78(11), pp.5670–5678. Yamauchi, Y., Daikoku, T., Goshima, F. & Nishiyama, Y., 2003. Herpes simplex virus UL14 protein blocks apoptosis. Microbiology and immunology, 47(9), pp.685–689. Yamauchi, Y., Kiriyama, K., Kubota, N., Kimura, H., Usukura, J. & Nishiyama, Y., 2008. The UL14 tegument protein of herpes simplex virus type 1 is required for efficient nuclear transport of the alpha transinducing factor VP16 and viral capsids. Journal of virology, 82(3),

- 219 -

References

pp.1094–106. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2224439&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Yamauchi, Y., Wada, K., Goshima, F., Daikoku, T., Ohtsuka, K. & Nishiyama, Y., 2002. Herpes simplex virus type 2 UL14 gene product has heat shock protein (HSP)-like functions. Journal of cell science, 115(Pt 12), pp.2517–2527. Yamauchi, Y., Wada, K., Goshima, F., Takakuwa, H., Daikoku, T., Yamada, M. & Nishiyama, Y., 2001. The UL14 protein of herpes simplex virus type 2 translocates the minor capsid protein VP26 and the DNA cleavage and packaging UL33 protein into the nucleus of coexpressing cells. Journal of General Virology, 82, pp.321–330. Yan, Q., Dutt, S., Xu, R., Graves, K., Juszczynski, P., Manis, J.P. & Shipp, M.A., 2009. BBAP Monoubiquitylates Histone H4 at Lysine 91 and Selectively Modulates the DNA Damage Response. Molecular Cell, 36(1), pp.110–120. Available at: http://dx.doi.org/10.1016/j.molcel.2009.08.019. Yan, Q., Xu, R., Zhu, L., Cheng, X., Wang, Z., Manis, J. & Shipp, M.A., 2013. BAL1 and its partner E3 ligase, BBAP, link Poly(ADP-ribose) activation, ubiquitylation, and double-strand DNA repair independent of ATM, MDC1, and RNF8. Molecular and cellular biology, 33(4), pp.845– 57. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3571337&tool=pmcentrez&rendertype=abstract. Yang, Q., Rout, M.P. & Akey, C.W., 1998. Three-dimensional architecture of the isolated yeast nuclear pore complex: functional and evolutionary implications. Molecular cell, 1(2), pp.223–234. Yatherajam, G., Huang, W. & Flint, S.J., 2011. Export of adenoviral late mRNA from the nucleus requires the Nxf1/Tap export receptor. Journal of virology, 85(4), pp.1429–1438. Yedowitz, J.C., Kotsakis, A., Schlegel, E.F.M. & Blaho, J. a, 2005. Nuclear localizations of the herpes simplex virus type 1 tegument proteins VP13/14, vhs, and VP16 precede VP22-dependent microtubule reorganization and VP22 nuclear import. Journal of virology, 79(8), pp.4730–4743. Yokoyama, N., Hayashi, N., Seki, T., Panté, N., Ohba, T., Nishii, K., Kuma, K., Hayashida, T., Miyata, T. & Aebi, U., 1995. A giant nucleopore protein that binds Ran/TC4. Nature, 376(6536), pp.184–188. Yu, C., Chen, Y., Lu, C. & Hwang, J., 2006. Prediction of Protein Subcellular Localization. Amino Acids, 651(December 2005), pp.643–651. Available at: http://www.ncbi.nlm.nih.gov/pubmed/16752418. Yu, S.W., Wang, H., Poitras, M.F., Coombs, C., Bowers, W.J., Federoff, H.J., Poirier, G.G., Dawson, T.M. & Dawson, V.L., 2002. Mediation of Poly ( ADP-Ribose ) Polymerase-1 – Dependent Cell Death by Apoptosis-Inducing Factor. Science, 297(5579), pp.259–63. Zachariae, U. & Grubmüller, H., 2008. Importin-β: Structural and Dynamic Determinants of a Molecular Spring. Structure. Zaichick, S. V, Bohannon, K.P., Hughes, A., Sollars, P.J., Pickard, G.E. & Smith, G.A., 2013. The Herpesvirus VP1/2 protein is an effector of dynein-mediated capsid transport and neuroinvasion. Cell Host Microbe, 13(2), pp.417–428. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3808164&tool=pmcentrez&rendertype=abstract [Accessed October 30, 2014]. Zaitseva, L., Cherepanov, P., Leyens, L., Wilson, S.J., Rasaiyaah, J. & Fassati, A., 2009. HIV-1 exploits importin 7 to maximize nuclear import of its DNA genome. Retrovirology, 6, p.11. Zhong, L. & Hayward, G.S., 1997. Assembly of complete, functionally active herpes simplex virus DNA replication compartments and recruitment of associated viral and cellular proteins in transient cotransfection assays. Journal of virology, 71(4), pp.3146–3160.

- 220 -

Appendix 1: List of primers

9 Appendix 1: List of primers

CCCAAGCTTACCATGACTCCCAAGAGGAAAAAGGGATTGGGCAAAGATAGCCCCCACAAGAAGCCCACCTCCGGCCGCCGCCTCCCTAGAT prTH1 CTGGG R4.EBV CCCAGATCTAGGGAGGCGGCGGCCGGAGGTGGGCTTCTTGTGGGGGCTATCTTTGCCCAATCCCTTTTTCCTCTTGGGAGTCATGGTAAGC prTH2 TTGGG R4.EBV complementary prTH3 CCCAAGCTTACCATGGATTGTCGTCGTAGACGCCGTCCTTCATGGACTCCTCCTTCAAGCGAGGAAAATTTAAGATCTGGG R4.VZV prTH4 CCCAGATCTTAAATTTTCCTCGCTTGAAGGAGGAGTCCATGAAGGACGGCGTCTACGACGACAATCCATGGTAAGCTTGGG R4.VZV complementary CCCAAGCTTACCATGTCACCCAAAAAAACACCCGAGAAACGCCGCAAGGACCTCAGCGGTAGCAAACACGGAGGCAAAAAGAAACCCAGAT prTH5 CTGGG R4.CMV CCCAGATCTGGGTTTCTTTTTGCCTCCGTGTTTGCTACCGCTGAGGTCCTTGCGGCGTTTCTCGGGTGTTTTTTTGGGTGACATGGTAAGC prTH6 TTGGG R4.CMV complementary CCCAAGCTTACCATGACGCGCCCGCGCAGACCTCGCGTCGTCATTCCTCCTTACGATCCGACAGACCGCCCACGACCGCCTCACCAAGACC prTH7 GCCCGAGATCTGGG R4.HHV-8 CCCAGATCTCGGGCGGTCTTGGTGAGGCGGTCGTGGGCGGTCTGTCGGATCGTAAGGAGGAATGACGACGCGAGGTCTGCGCGGGCGCGTC prTH8 ATGGTAAGCTTGGG R4.HHV-8 complementary UL35 PCR primer (antisense strand of prTH9 CTACATCACGTGCGCATG VP1-2 C-terminus) UL35 PCR primer (sense strand of C- prTH10 GTGGGTGGTTGGTGCTG terminus of UL34) β-gal sequencing primer N-terminus prTH11 CATTCGCCATTCAGGCTG reverse prTH12 GGATTGGGCAAAGATAG β-gal.EBV PCR prTH13 CACCCGAGAAACGCCGC β-gal.CMV PCR prTH14 GTCGTCATTCCTCCTTAC β-gal.HHV-8 PCR prTH15 GATTGTCGTCGTAGACG β-gal.VZV PCR prTH16 GAAAGCCGATCCCAAAC V5 PCR primer prTH17 [Phos]AGCTTGCCACCATGGGAAAGCCGATCCCAAACCCTTTGCTGGGATTGGACTCCACCA V5 tag forward prTH18 [Phos]AGCTTGGTGGAGTCCAATCCCAGCAAAGGGTTTGGGATCGGCTTTCCCATGGTGGCA V5 tag reverse

nested primer to amplify VP26 region; prTH19 GCCTATAAAAAAGGACGCACCG goes with primers prTH9 and TH10

- 221 -

Appendix 1: List of primers

nested primer to amplify VP26 region; prTH20 CCTATCCCACTCCCTTGAATAAACA goes with primers prTH9 and TH10 prTH21 gggaagcttaccatgcttcgcaagcgccgccgacccacctggactccgccttccagcgtcgaagacctgagatctccc R4.HSV P427R prTH22 gggagatctcaggtcttcgacgctggaaggcggagtccaggtgggtcggcggcgcttgcgaagcatggtaagcttccc R4.HSV P427R complementary prTH23 gggaagcttaccatgcttcccaggcgccgccgacccacctggactccgccttccagcgtcgaagacctgagatctccc R4.HSV K428R prTH24 gggagatctcaggtcttcgacgctggaaggcggagtccaggtgggtcggcggcgcctgggaagcatggtaagcttccc R4.HSV K428R complementary prTH25 gggaagcttaccatgtgtcgtcgtagacgccgtccttcatggactcctccttcaagcgaggaaaatttaagatctccc R4.VZV (w/o leader aspartate) prTH26 gggagatcttaaattttcctcgcttgaaggaggagtccatgaaggacggcgtctacgacgacacatggtaagcttccc R4.VZV (no D) complementary prTH27 gggaagcttaccatgtgtcgtaagagacgccgtccttcatggactcctccttcaagcgaggaaaatttaagatctccc R4.VZV (no D); R2>K prTH28 gggagatcttaaattttcctcgcttgaaggaggagtccatgaaggacggcgtctcttacgacacatggtaagcttccc R4.VZV (no D); R2>K; complementary prTH29 gggaagcttaccatgtgtcctcgtagacgccgtccttcatggactcctccttcaagcgaggaaaatttaagatctccc R4.VZV (no D); R1>P prTH30 gggagatcttaaattttcctcgcttgaaggaggagtccatgaaggacggcgtctacgaggacacatggtaagcttccc R4.VZV (no D); R1>P; complementary prTH31 gggaagcttaccatgtgtcctaagagacgccgtccttcatggactcctccttcaagcgaggaaaatttaagatctccc R4.VZV (no D); R1>P; R2>K

R4.VZV (no D); R1>P; R2>K; prTH32 gggagatcttaaattttcctcgcttgaaggaggagtccatgaaggacggcgtctcttaggacacatggtaagcttccc complementary CCCAAGCTTACCATGGGAAACAAAGGACGCGGCGGTAACAAAGGACGCGGCGGAAAGACGGGACGTGGCGGAAATGAAGGACGCGGTAGAT prTH33 CTGGG R5.HHV-8 CCCAGATCTACCGCGTCCTTCATTTCCGCCACGTCCCGTCTTTCCGCCGCGTCCTTTGTTACCGCCGCGTCCTTTGTTTCCCATGGTAAGC prTH34 TTGGG R5.HHV-8 complementary prTH35 CCCAAGCTTACCATGCTAACTGCCACAAGAGGGCAGAAACGCAAATTTTCCTCGCTTAGATCTGGG R9.HHV-8 prTH36 CCCAGATCTAAGCGAGGAAAATTTGCGTTTCTGCCCTCTTGTGGCAGTTAGCATGGTAAGCTTGGG R9.HHV-8 complementary

forward primer to clone VZV homolog of prTH37 CCCGGATCCGATATAATTCCGCCTATAGCTGTC UL36 with BamHI site reverse primer to clone VZV homolog of prTH38 CCCGGATCCGCTACTTTATATATATGTTCC UL36 cloning RFP with N-terminal BamHI and prTH39 GGGGGATCCGCTCTGTCAAAGCACGG C-terminal XbaI site

- 222 -

Appendix 1: List of primers

cloning RFP with N-terminal BamHI and prTH40 GGGTCTAGATTATCCGGGCAATGCGG C-terminal XbaI site HA tag with HindIII on both sides; 2nd prTH41 [Phos]AGCTTGCCACCATGTACCCATACGATGTTCCAGATTACGCTC HindIII lost upon integration HA tag with HindIII on both sides; 2nd prTH42 [Phos]AGCTGAGCGTAATCTGGAACATCGTATGGGTACATGGTGGCA HindIII lost upon integration V5 tag with HindIII on both sides; 2nd prTH43 [Phos]AGCTTGCCACCATGGGAAAGCCGATCCCAAACCCTTTGCTGGGATTGGACTCCACCC HindIII lost upon integration V5 tag with HindIII on both sides; 2nd prTH44 [Phos]AGCTGGGTGGAGTCCAATCCCAGCAAAGGGTTTGGGATCGGCTTTCCCATGGTGGCA HindIII lost upon integration forward primer to confirm insertion of prTH45 GAAAGCCGATCCCAAAC V5 tag forward primer to clone EGFP-C2 with prTH46 ccctctagaggtttagtgaaccgtcagatccgc N-terminal XbaI site reverse primer to clone EGFP-C2 with C- prTH47 gcgccatagatccggtggagctcgagatctgagtccggcc terminal XcmI site alternative forward primer to prTH48 CCCAAGCTTGCCACCATGGTGAGCAAGGG complement prTH47 (+ 5' HindIII site) forward primer to confirm insertion of prTH49 GTACCCATACGATGTTCCAG HA tag antisense primer. anneals perfectly to prTH52 AGGTGGGTCGGCGGCGAGC K>A mutation (last three nucleotides are complementary to the mutated codon) sense primer to amplify region 1188bp- prTH53 GGACGAAGCGTTCAGCGAAC 2351bp of NT6 (BamHI-MluI) anti-sense primer to amplify region prTH54 GAGTTTTCGTCCAGCACGTGC 1188bp-2351bp of NT6 (BamHI-MluI) prTH55 GGTCTTGGTGAGGCGGTCGT goes with prTH53 prTH56 ACGACCGCCTCACCAAGACC goes with prTH54 prTH57 cagaggcggccgaagaagatg BamHI-MluI sequencing prTH58 ggtgccgaggcgaggacag BamHI-MluI sequencing primer in NT6v3.CMV (antisense) to prTH59 cgtgtttgctaccgctgagg confirm insertion. goes with prTH53 primer in NT6v3.CMV (sense) to confirm prTH60 cctcagcggtagcaaacacg insertion. goes with prTH54 sense primer within MyoPondin NLS prTH63 ccggttggtaagaagaggagg insert; paired with prTH54

- 223 -

Appendix 1: List of primers

sense primer in mutR5 to confirm virus; prTH64 ggaggcaacggcattgtcg paired with prTH54 sense primer to confirm wild type NLS in prTH65 gccgacccacctggactc virus; paired with prTH54 sense primer to confirm deletion of NLS prTH66 ctactcggccggcacctg in virus; paired with prTH54

prTH67 CCCGCTAGCGCCACCATGGGTAAACCAATTCCTAATCCTCTACTTGGCCTAGATAGTACGAGCGGCCGCTGG V5-tag flanked by NheI/NotI; sense oligo

V5-tag flanked by NheI/NotI; antisense prTH68 CCAGCGGCCGCTCGTACTATCTAGGCCAAGTAGAGGATTAGGAATTGGTTTACCCATGGTGGCGCTAGCGGG oligo sense primer to go with pTH54; anneals prTH73 ctgacttcgggggagaaaacga to the wild type R5 to confirm second basic patch made by Beckman Genomics; sense prTH74 ccaatgtgcctggatgcgttc primer to C-terminal region of GST approx. 200bp away from MCS; sense oligo; R1.HSV-1 region with prTH85 tcgagatgggcgtcgtcccggttggtaagaagaggaggaggagggccaggaagacccccgtcggggccgcggtccccg Myopodin NLS instead of R4/R5 antisense oligo; R1.HSV-1 region with prTH86 ctagcggggaccgcggccccgacgggggtcttcctggccctcctcctcctcttcttaccaaccgggacgacgcccatc Myopodin NLS instead of R4/R5

tcgagatgggcgtcgtcccggttggtaaacgcccggcggcgaccaagaaagcgggccaggcgaagaagaagaaaacccccgtcggggccgc sense oligo; R1.HSV-1 region with prTH87 ggtccccg Nucleoplasmin NLS instead of R4/R5

ctagcggggaccgcggccccgacgggggttttcttcttcttcgcctggcccgctttcttggtcgccgccgggcgtttaccaaccgggacga antisense oligo; R1.HSV-1 region with prTH88 cgcccatc Nucleoplasmin NLS instead of R4/R5

sense primer to 5'-end of pTRE-tight prTH89 gtagcggatcccgagtttactccctatcagtg operator/promoter; incl. 5' BamHI site antisense primer to 3'-end of pTRE-tight prTH90a/b gtacgctcgagctccaggcgatctgacggttc operator/promoter; incl. 3' XhoI site antisense primer to 3'-end of pTRE-tight CGATGAATTCAATGATCAGATCTAGAGCAGATCTCCATCGATATAGATATCTCACGCGTTGAATGCGGCCGCTTGAATCTCGAGCTCCAGG prTH90c CGATCTGACGGTTC operator/promoter; incl. Multi cloning site V5 tag at N-terminus of NT3; including prTH93 GTGGCTAGCGCCACCATGGGAAAGCCGATCCCAAACC Kozak prTH94 CCAGTGTGATGGATATCTGCAG NT3 antisense primer; with prTH93

- 224 -

Appendix 1: List of primers

incl. 5' HindIII, followed by Kozak and prTH95a Gcgaagcttgccaccatgtcccctatactaggttattgg GST prTH96a gaattcaagcttggatccaggggcccctggaacag for cloning from pCAG-GST-NT4 vector 60-mer dsDNA oligo complementary to prTH98; used to induce a STING prTH97 TAAGACACGATGCGATAAAATCTGTTTGTAAAATTTATTAAGGGTACAAATTGCCCTAGC dependent response (probably IFI16 initiated) 60-mer dsDNA oligo complementary to prTH97; used to induce a STING prTH98 GCTAGGGCAATTTGTACCCTTAATAAATTTTACAAACAGATTTTATCGCATCGTGTCTTA dependent response (probably IFI16 initiated) sense primer to Neomycin C-terminal prTH99 ctggattcatcgactgtg domain; used to confirm pQ-TetR- Neo_v2 antisense primer to confirm pQ-TetR- prTH100 gctggtgatattgttgagtc Neo_v2 Forward primer to clone BamHI-MluI prTH101 gtgcacaattggcggatcccccctccgccgac fragment from pTH19 into pCAG-GST, includes MfeI site (lost upon cloning) Reverse primer to clone BamHI-MluI prTH102 GAGCTGAATTCGAGGCCCAGCCCGAGCTC fragment from pTH19 into pCAG-GST, includes EcoRI site 5' forward primer for amplification of TetON (rtTA) from Clontech vector; NotI prTH103a Gatcagcggccgcccgaattcatatgtctaga included upstream of ORF; using this method the HA-tag of pQ-HA will be lost 5' forward primer for amplification of TetON (rtTA) from Clontech vector; AgeI included upstream of ORF; lacks first prTH103b Gatcaaccggttctagattagataaaagtaaag methionine using this method the HA- tag of pQ-HA will be kept giving rise to a HA-rtTA fusion 3' reverse primer for amplification of rtTA; goes with prTH103a/b; use prTH104 cttatcatgtctggatcctc naturally occurring BamHI site for cloning REV NES in frame into XhoI-NheI inserts prTH105 cacgctagcttgccgcctcttgagagattgaccctaagatctgac via NheI-BglII. For transfer into EGFP-C3 in frame.

- 225 -

Appendix 1: List of primers

REV NES complementary in frame into prTH106 gtcagatcttagggtcaatctctcaagaggcggcaagctagcgtg XhoI-NheI inserts via NheI-BglII. For transfer into EGFP-C3 in frame. reverse primer to confirm insertion of prTH107 GGGTCAATCTCTCAAGAGG the REV NES sequence into pTH29 or any other. pair with HA-primer forward primer to pTH117/118; cloning prTH108 gatctacgcgtgaattcGccaccatggtgagcaag of EGFP-NLS-NES or GFP-NLS into pQ- Flag or pQ-CMV-MCS-IRES-Neo_v2 reverse primer to pTH117/118; cloning prTH109 gctaagcggccgcgaattccagttatctagatccggtgg of EGFP-NLS-NES or GFP-NLS into pQ- Flag or pQ-CMV-MCS-IRES-Neo_v2

- 226 -

Appendix 2: Alignment of R1 regions of HSV-1 strains

10 Appendix 2: Alignment of R1 regions of HSV-1 strains

- 227 -

Appendix 3: MS results for excised bands from NT3 IP

11 Appendix 3: MS results for excised bands from NT3 IP

Intensity Unique peptides PEP MS/MS MW Ids Protein name Gene name NT3 NT3NLS score Count NT3 NT3NLS

Q5JP53 P07437 Q5ST81 1 24,5413 0 0 207 2 0 48 Tubulin beta chain TUBB E9PBJ4

2E- P68363 Q9BQE3 F5H5D3 Tubulin alpha-1B chain, Tubulin alpha-1C 2 27,5482 0 158 6 1 50 TUBA1B TUBA1C TUBA4A 280 A8MUB1 P68366 F8VVB9 chain, Tubulin alpha-4A chain

9E- O00571 B4DXX7 O15523 ATP-dependent RNA helicase DDX3X, ATP- 3 28,9973 26,9931 112 10 8 73 DDX3X DDX3Y 261 B4E3C4 B4DLA0 C9J8G5 dependent RNA helicase DDX3Y

4 25,8649 0 0 105 2 0 113 P09874 Q5VX84 Q5VX85 Poly [ADP-ribose] polymerase 1 PARP1

8E- Q92841 H3BLZ8 C9JMU5 Probable ATP-dependent RNA helicase 5 27,2874 24,1357 75 9 4 80 DDX17 174 G5E9L5 DDX17

3E- P11940 E7ERJ7 E7EQV3 6 26,6509 0 67 3 1 71 Polyadenylate-binding protein 1 PABPC1 168 H0YAR2 H0YBN4 Q9H361

ADP/ATP translocase 2, ADP/ATP 7 24,2034 0 2E-65 66 2 0 33 P05141 SLC25A5 translocase 2, N-terminally processed 6E- 8 25,6427 0 52 4 1 81 P33993 C9J8M6 DNA replication licensing factor MCM7 MCM7 175 4E- Vacuolar protein sorting-associated 9 30,9121 28,7984 46 13 12 92 Q96QK1 I3L4S0 I3L4P4 VPS35 209 protein 35 6E- 10 26,5806 0 34 4 1 84 Q8TDB6 E3 ubiquitin-protein ligase DTX3L DTX3L 174

8E- P62258 K7EM20 B4DJF2 11 27,5987 0 32 3 0 29 14-3-3 protein epsilon YWHAE 103 I3L3T1 K7EIT4 I3L0W5

P46060 F8W7I9 H0Y4Q3 12 26,0217 0 4E-86 28 2 1 64 Ran GTPase-activating protein 1 RANGAP1 B0QYT5 B0QYT4 B0QYT6

P31947 E5RGE1 Q4VY20 13 26,2397 0 2E-09 4 2 0 28 14-3-3 protein sigma SFN E5RIR4 E9PD24 E7EVZ2

- 228 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

Selected proteins identified by MS in SDS-PA gel slices from NT3 and NT3ΔNLS sample lanes. Details about the experiment can be found in Figure 23. Bands demarcated with black arrowheads were excised, proteins isolated and tryptically digested. Peptides were analysed by MS and the 13 most interesting proteins that were at least four fold enriched in the NT3 slices compared to the NT3ΔNLS slices are listed. Relative amount of a protein is given on a log2 scale, that is a change of 1 unit equals a doubling in abundance.

12 Appendix 4: Category III (proteins at least four fold enriched versus control)

Intensity MS/MS unique peptides functional PEP score MW Protein names Gene names localisation NT3 NT3 ΔNLS paren tal Count NT3 NT3NLS control annotation DNA replication licensing factor 1 30,0184 27,6219 27,1171 5,68E-175 52 26 18 18 81 MCM7 cell cycle Nuclear MCM7

2 32,3941 29,2294 29,2263 3,18E-120 43 13 9 8 22 60S ribosomal protein L18 RPL18 translation Cytoplasmic

3 31,6468 29,5047 29,2383 3,20E-45 42 13 9 8 13 60S ribosomal protein L26 RPL26 KRBA2 translation Cytoplasmic

E3 ubiquitin-protein ligase ubiquitin and Nuclear, 4 30,0753 27,4888 0 6,27E-174 34 18 15 0 84 DTX3L DTX3L SUMO pathway cytoplasmic RNA polymerase II-associated 5 28,3236 26,2524 26,1291 3,37E-63 17 12 5 8 60 PAF1 transcription nuclear factor 1 homolog Mediator of RNA polymerase II 6 28,3477 24,2055 23,0033 6,29E-86 14 15 4 3 161 MED14 transcription Nuclear transcription subunit 14 Testis-specific Y-encoded-like 7 27,6883 25,3861 25,6694 6,79E-43 11 7 4 5 49 TSPYL1 genome structure nuclear protein 1 tRNA-splicing endonuclease translation, tRNA 8 27,2093 0 0 1,14E-32 7 7 1 1 34 TSEN34 nuclear subunit Sen34 metabolism Leucine-rich repeat-containing ubiquitin and 9 27,0931 24,7393 24,7187 1,61E-31 7 5 4 3 89 LRRC41 membrane protein 41 SUMO pathway MKI67 FHA domain-interacting 10 25,6633 0 0 7,02E-25 6 5 1 0 34 MKI67IP RNA metabolism nuclear nucleolar phosphoprotein Actin, alpha skeletal muscle Actin, alpha cardiac muscle 1 ACTA1 ACTC1 11 25,974 0 0 6,63E-55 6 2 1 1 42 Actin, gamma-enteric smooth cytoskeletal cytoplasm ACTG2 ACTA2 muscle Actin, aortic smooth muscle

- 229 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

SWI/SNF-related matrix- associated actin-dependent 12 25,1609 0 0 6,99E-43 5 3 2 0 58 SMARCD1 genome structure nuclear regulator of chromatin subfamily D member 1 Uncharacterized protein 13 25,7742 0 23,452 8,36E-16 5 5 1 3 98 CXorf57 not annotated not annotated CXorf57 Histone acetyltransferase type nuclear, 14 26,2469 0 0 3,02E-27 4 2 1 1 50 HAT1 genome structure B catalytic subunit cytoplasmic Testis-specific Y-encoded-like 15 25,4644 22,9001 22,9826 1,48E-19 4 3 2 2 79 TSPYL2 transcription nuclear protein 2 Mediator of RNA polymerase II 16 24,78 22,6472 22,0634 2,95E-14 4 3 3 2 110 MED24 transcription Nuclear transcription subunit 24 protein kinase, cell 17 25,5797 0 0 2,08E-13 4 4 2 2 20 Aurora kinase B AURKB nuclear cycle Putative methyltransferase 18 24,1387 0 0 2,96E-12 4 3 1 1 47 NSUN5 RNA metabolism nuclear NSUN5 Exosome complex component nuclear, 19 24,3653 0 0 2,09E-10 4 2 1 1 30 EXOSC8 RNA metabolism RRP43 cytoplasmic Leucine-rich repeat and WD nuclear, 20 24,2364 0 0 8,23E-08 4 2 1 2 71 LRWD1 cell cycle repeat-containing protein 1 cytoplasmic 21 23,7259 0 0 1,91E-07 4 2 1 1 21 Cytochrome b-245 light chain CYBA metabolism membrane Transcription initiation factor TFIID subunit 9 Transcription 22 24,03 0 0 2,74E-07 4 3 0 0 22 TAF9 TAF9B transcription nuclear initiation factor TFIID subunit 9B 23 24,3195 0 0 4,539E-07 4 3 1 0 77 Cirhin CIRH1A transcription nuclear p21-activated protein kinase- 24 24,1941 0 0 9,69E-07 4 2 1 1 44 PAK1IP1 cell cycle nuclear interacting protein 1 RNA metabolism, 25 23,1428 0 0 3,202E-09 4 2 2 2 219 Endoribonuclease Dicer DICER1 cytoplasmic miRNA biogenesis 26 25,1453 0 0 6,05E-09 4 4 1 2 309 Protein PRRC2C PRRC2C mRNA metabolism Cytoplasmic nuclear, 27 24,8918 0 0 3,57E-05 3 2 0 0 15 Galectin-7 LGALS7 apoptosis cytoplasmic, exosomes protein kinase, multiple Nuclear, 28 26,4832 24,3834 0 4,56E-14 3 3 2 1 25 Casein kinase II subunit beta CSNK2B pathways, cytoplasmic differentiation nuclear, 29 24,9509 0 0 1,08E-13 3 2 1 1 19 Gem-associated protein 6 GEMIN6 mRNA metabolism cytoplasmic

- 230 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

Non-structural maintenance of genome structure, 30 26,2117 0 0 1,60E-30 3 3 1 1 44 chromosomes element 4 NSMCE4A nuclear DNA repair homolog A Mediator of RNA polymerase II 31 25,749 0 0 1,51E-18 3 5 2 2 97 MED16 transcription nuclear transcription subunit 16 nuclear, 32 24,7728 0 0 7,43E-11 3 2 1 1 17 M-phase phosphoprotein 6 MPHOSPH6 RNA metabolism cytoplasmic RNA-binding protein 27 RNA- RBM27 nuclear, 33 23,9499 0 0 2,107E-09 3 2 1 1 115 mRNA metabolism binding protein 26 RBM26 cytoplasmic General transcription factor 3C 34 25,7301 0 0 5,54E-13 3 3 1 1 101 GTF3C2 transcription nuclear polypeptide 2 ubiquitin and nuclear, 35 24,6446 0 0 2,92E-09 3 3 0 0 69 E3 ubiquitin-protein ligase RLIM RLIM SUMO pathway cytoplasmic Origin recognition complex 36 24,1607 0 0 3,60E-12 3 3 1 1 66 ORC2 cell cycle nuclear subunit 2 DNA polymerase delta subunit 37 25,459 0 0 4,61E-12 3 3 1 1 50 POLD2 cell cycle nuclear 2 39S ribosomal protein L53, 38 24,1682 0 0 5,21E-12 3 2 1 1 12 MRPL53 translation mitochondrium mitochondrial 39 24,4644 0 0 1,25E-11 3 3 2 3 106 WD repeat-containing protein 3 WDR3 RNA metabolism nuclear Leucine-rich repeat flightless- nuclear, 40 23,922 0 0 4,22E-09 3 3 1 3 89 LRRFIP1 transcription interacting protein 1 cytoplasmic CUGBP Elav-like family member nuclear, 41 23,8516 0 0 1,98E-08 3 3 1 1 52 CELF1 mRNA metabolism 1 cytoplasmic Prostate tumor-overexpressed membrane, 42 23,5272 0 0 1,87E-10 3 2 0 1 43 PTOV1 transcription gene 1 protein nuclear Zinc finger CCCH domain- 43 23,7276 0 0 5,125E-10 3 2 1 1 106 ZC3H18 mRNA metabolism nuclear containing protein 18 nuclear, 44 24,6918 0 0 2,12E-05 3 2 1 1 27 Protein lin-28 homolog B LIN28B RNA metabolism cytoplasmic Protein strawberry notch SBNO1 45 23,4686 0 0 1,57E-07 3 2 1 2 154 homolog 1 Protein strawberry transcription nuclear SBNO2 notch homolog 2 Ribosome biogenesis protein 46 24,6081 0 0 1,09E-06 3 3 1 1 48 WDR12 RNA metabolism nuclear WDR12 47 24,4627 0 0 2,90E-06 3 2 1 1 12 C11orf48 not annotated not annotated Transcriptional activator protein Pur-beta 48 24,0945 0 0 1,86E-04 3 2 1 1 33 PURB PURA transcription nuclear Transcriptional activator protein Pur-alpha

- 231 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

Cytoplasmic, 49 24,6364 0 0 2,76E-12 3 4 3 2 188 Golgin subfamily B member 1 GOLGB1 golgi Mitotic spindle assembly 50 25,2247 0 0 1,70E-10 3 3 2 1 24 MAD2L1 cell cycle cytoplasmic checkpoint protein MAD2A p53 and DNA damage- 51 24,1257 0 0 5,28E-09 3 2 1 1 16 PDRG1 DNA repair cytoplasmic regulated protein 1 MAP7 domain-containing 52 25,6583 0 0 1,05E-08 3 2 1 1 93 MAP7D1 cytoskeletal cytoplasm protein 1 nuclear, 53 23,8272 0 0 8,406E-06 2 2 0 0 75 Symplekin SYMPK mRNA metabolism membrane, cytoplasmic 54 24,9167 0 0 1,07E-07 2 2 1 0 26 Transmembrane protein 109 TMEM109 apoptosis nuclear, ER Cyclin-dependent kinase 2 protein kinase, cell Nuclear, 55 25,0132 0 0 3,92E-13 2 2 0 0 27 CDK2 CDK3 Cyclin-dependent kinase 3 cycle cytoplasmic nuclear, 56 23,5498 0 0 1,27E-07 2 2 0 0 22 Kinetochore protein Spc24 SPC24 cell cycle cytoplasmic SUMO-activating enzyme ubiquitin and nuclear, 57 25,3824 0 0 1,46E-07 2 3 1 1 71 UBA2 subunit 2 SUMO pathway cytoplasmic transcription, Helicase-like transcription nuclear, 58 23,6703 0 0 1,80E-05 2 2 0 0 48 HLTF ubiquitin and factor cytoplasmic SUMO pathway

Histone H2A.V Histone H2A.Z H2AFV H2AFZ 59 25,4414 0 0 4,38E-19 2 2 0 1 14 Histone H2A Histone H2A type genome structure nuclear HIST1H2AA 1-A

Ras GTPase-activating-like 60 24,3738 0 0 2,69E-15 2 2 1 1 179 IQGAP3 small GTPase membrane protein IQGAP3 General transcription factor IIH 61 24,618 0 22,546 6,17E-14 2 2 1 2 16 GTF2H3 transcription nuclear subunit 3 62 24,7109 0 0 6,33E-13 2 4 1 0 100 Pre-mRNA-splicing factor SYF1 XAB2 mRNA metabolism nuclear U3 small nucleolar 63 23,6613 0 0 6,34E-13 2 2 1 0 22 IMP3 RNA metabolism nuclear ribonucleoprotein protein IMP3 G patch domain-containing 64 24,2329 0 0 1,31E-12 2 2 0 1 14 GPATCH4 mRNA metabolism nuclear protein 4 Ribosome biogenesis protein 65 24,2898 0 0 3,53E-12 2 3 2 2 146 BMS1 translation nuclear BMS1 homolog Cyclin-dependent kinase 66 26,633 0 0 5,10E-11 2 2 1 1 14 CDKN2A cell cycle nuclear inhibitor 2A, isoform 4

- 232 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

DNA polymerase DNA 67 24,9746 0 0 7,92E-10 2 3 1 1 258 polymerase epsilon catalytic POLE cell cycle nuclear subunit A N-acylneuraminate nucleotide nuclear, 68 25,3114 0 0 5,11E-09 2 2 0 1 48 CMAS cytidylyltransferase metabolism membrane Ribosome production factor 2 translation, mRNA 69 24,6684 0 0 8,03E-09 2 2 1 1 25 RPF2 nuclear homolog metabolism DNA-directed RNA polymerase II subunit RPB11-b2 DNA- POLR2J3 directed RNA polymerase II POLR2J2 70 24,907 0 0 3,13E-08 2 2 1 1 13 transcription nuclear subunit RPB11-b1 DNA-directed POLR2J RNA polymerase II subunit POLR2J1 RPB11-a cytoplasmic, 71 22,8072 0 0 9,19E-04 2 2 1 0 40 Modulator of apoptosis 1 MOAP1 apoptosis mitochondrium 72 22,2295 0 0 2,13E-07 2 2 0 0 89 Protein FAM91A1 FAM91A1 not annotated not annotated Zinc finger and BTB domain- 73 24,0277 0 0 1,389E-06 2 2 1 0 119 ZBTB11 transcription nuclear containing protein 11

Heterogeneous nuclear hCG_2044799 74 24,0888 0 0 9,01E-06 2 2 1 0 85 ribonucleoprotein U-like mRNA metabolism nuclear HNRNPUL2 protein 2

Ribonuclease P/MRP protein 75 22,7283 0 0 9,118E-06 2 2 1 1 19 POP5 RNA metabolism nuclear subunit POP5 76 22,7954 0 0 3,82E-05 2 2 0 0 113 Genetic suppressor element 1 GSE1 not annotated not annotated 77 23,2631 0 0 4,39E-05 2 2 1 1 24 RNA-binding protein 34 RBM34 RNA binding nuclear Cell division cycle 7-related protein kinase, cell 78 23,4769 0 0 3,10E-04 2 2 0 0 61 CDC7 nuclear protein kinase cycle Zinc finger protein 225 Zinc ZNF225 79 26,3113 0 0 4,26E-04 2 2 1 1 82 finger protein 221 Zinc finger ZNF221 transcription nuclear protein 224 ZNF224 membrane, 80 26,1687 0 0 8,64E-04 2 2 0 1 104 Exocyst complex component 2 EXOC2 exocytosis golgi E3 ubiquitin-protein ligase ubiquitin and 81 31,2931 0 0 2,06E-06 2 2 1 1 201 LTN1 cytoplasmic listerin SUMO pathway Receptor expression-enhancing 82 24,3401 0 0 9,87E-06 2 2 0 0 22 REEP4 cell cycle Cytoplasmic protein 4 Regulator of nonsense 83 22,9113 0 0 1,49E-03 2 2 1 1 148 UPF2 RNA metabolism cytoplasmic transcripts 2 84 23,7052 0 0 0,0015373 2 2 1 1 134 Sperm-associated antigen 5 SPAG5 cell cycle cytoplasmic

- 233 -

Appendix 4: Category III (proteins at least four fold enriched versus control)

Selenocysteine-specific nuclear, 85 24,826 0 0 5,52E-11 1 3 2 2 65 EEFSEC translation elongation factor cytoplasmic nuclear, 86 24,0317 0 0 2,86E-10 1 2 1 1 63 Protein AATF AATF apoptosis cytoplasmic Cyclin-dependent kinase inhibitor 2A, isoforms 1/2/3 CDKN2A nuclear, 87 23,5443 0 0 1,704E-06 1 2 1 0 11 cell cycle Cyclin-dependent kinase 4 CDKN2B cytoplasmic inhibitor B amino acid Diphthamide biosynthesis nuclear, 88 22,8387 0 0 4,29E-04 1 2 0 1 47 DPH1 metabolism and protein 1 cytoplasmic modification Neuroepithelial cell- small GTPase, cell cytoplasmic, 89 24,3684 0 0 5,42E-06 1 2 0 1 68 NET1 transforming gene 1 protein cycle membrane

90 25,7769 0 0 1,31E-13 1 2 1 0 48 Histone-binding protein RBBP4 RBBP4 genome structure nuclear

Macrophage erythroblast 91 24,2344 0 0 8,64E-10 1 3 1 2 21 MAEA apoptosis membrane attacher TFIIH basal transcription factor 92 23,923 0 0 1,92E-08 1 2 1 1 81 ERCC2 transcription nuclear complex helicase XPD subunit 93 24,5032 0 0 1,68E-07 1 3 1 1 65 Zinc finger protein 512 ZNF512 transcription nuclear 94 23,2994 0 0 3,649E-07 1 2 1 1 93 Zinc finger protein 598 ZNF598 mRNA metabolism nuclear 28S ribosomal protein S18b, 95 24,4966 0 0 6,65E-07 1 2 1 0 29 MRPS18B translation mitochondrium mitochondrial 96 24,9449 0 0 3,40E-06 1 2 0 1 35 Actin-related protein 2 ACTR2 cytoskeletal cytoplasmic

97 23,4385 0 0 3,46E-05 1 2 1 1 616 Dystonin DST cytoskeletal Cytoplasmic

Dimethyladenosine transferase 98 24,2085 0 0 8,75E-05 1 2 1 1 45 TFB2M transcription mitochondrium 2, mitochondrial Splicing factor, arginine/serine- transcription, 99 24,2223 0 0 1,33E-03 0 2 1 1 139 SCAF1 nuclear rich 19 splicing

- 234 -

Appendix 5: Category IV (proteins at least four fold enriched)

13 Appendix 5: Category IV (proteins at least four fold enriched)

Intensity unique peptides MS/MS Gene functional paren PEP score MW Protein names localisation NT3 NT3 ΔNLS Count NT3 NT3NLS control names annotation tal cytoplasmic, Very long-chain specific acyl-CoA 1 0 26,7455 24,2334 1,04E-50 7 0 8 2 70 ACADVL metabolism nuclear, dehydrogenase, mitochondrial mitochondrium cytoplasmic, ER, 2 0 26,6598 0 9,31E-14 7 1 4 0 11 Protein S100-A7 S100A7 immunity exosomes mRNA 3 24,341 26,3518 24,2033 1,97E-16 5 2 4 2 21 Nucleolar protein 16 NOP16 nuclear metabolism 4 0 24,019 0 2,25E-09 4 1 2 2 34 Estradiol 17-beta-dehydrogenase 12 HSD17B12 metabolism ER apoptosis, cytoplasmic, 5 0 24,6258 0 9,18E-17 3 2 2 1 24 Zinc finger protein ubi-d4 DPF2 transcription nuclear Mitochondrial import inner membrane 6 0 25,3732 0 1,48E-15 3 2 2 1 11 TIMM13 chaperone mitochondrium translocase subunit Tim13 7 0 23,8669 0 1,73E-10 3 2 2 2 72 HAUS augmin-like complex subunit 5 HAUS5 cell cycle cytoplasmic 8 0 24,1955 0 3,26E-08 3 1 3 1 24 Protein THEM6 THEM6 not annotated secreted Beta-hexosaminidase Beta-hexosaminidase 9 0 23,2888 0 3,06E-07 3 1 2 2 58 HEXA metabolism lysosome subunit alpha Mitochondrial tRNA-specific 2-thiouridylase RNA 10 0 24,1923 0 4,82E-07 3 0 2 1 48 TRMU mitochondrium 1 metabolism Prostaglandin E synthase 2 Prostaglandin E metabolism, 11 0 24,275 0 2,69E-06 3 0 3 1 42 PTGES2 cytoplasmic, Golgi synthase 2 truncated form immunity Serine/threonine-protein phosphatase 4 regulatory subunit 3A Serine/threonine- SMEK1 cytoplasmic, 12 0 23,8981 0 2,76E-05 3 1 2 1 62 DNA repair protein phosphatase 4 regulatory subunit SMEK2 nuclear 3B 13 0 23,6613 0 9,15E-17 2 1 2 1 10 Dynein light chain Tctex-type 1 DYNLT1 cytoskeletal cytoplasmic, Golgi RNA 14 0 25,7572 23,741 4,90E-13 2 1 2 2 42 Ribonuclease P protein subunit p40 RPP40 nuclear metabolism

cytoplasmic, 15 0 24,2274 0 1,76E-12 2 1 3 1 76 THO complex subunit 1 THOC1 transcription nuclear

16 0 22,8592 0 5,72E-08 2 1 2 1 44 Metastasis-associated protein MTA3 MTA3 transcription nuclear 17 0 25,879 0 1,35E-07 2 2 3 2 43 GTP-binding protein 10 GTPBP10 translation nuclear

- 235 -

Appendix 5: Category IV (proteins at least four fold enriched)

Intensity unique peptides MS/MS Gene functional paren PEP score MW Protein names localisation NT3 NT3 ΔNLS Count NT3 NT3NLS control names annotation tal 18 0 24,0409 0 1,06E-06 2 1 2 1 19 Replication protein A 32 kDa subunit RPA2 cell cycle nuclear 19 0 24,4321 0 2,64E-06 2 1 2 0 285 Spectrin alpha chain, non-erythrocytic 1 SPTAN1 cytoskeletal cytoplasmic GNAS Guanine nucleotide-binding protein G(s) cytoplasmic, 20 0 25,4809 0 7,06E-06 2 2 2 1 46 GNAT1 signalling subunit alpha isoforms short membrane GNAI1 21 0 24,2396 0 1,12E-05 2 1 3 2 120 Importin-8 IPO8 nuclear import cytoplasmic Cytochrome c oxidase assembly protein 3 22 0 24,7104 0 2,19E-05 2 1 2 1 8 COA3 translation mitochondrium homolog, mitochondrial 23 0 23,1143 0 6,57E-05 2 1 2 1 46 Nuclear pore glycoprotein p62 NUP62 nuclear pore cytoplasm, nuclear 24 0 24,9024 0 6,90E-05 2 2 2 1 15 39S ribosomal protein L41, mitochondrial MRPL41 translation mitochondrium 25 0 22,5174 0 0,0002895 2 1 2 1 41 NADPH--cytochrome P450 reductase POR metabolism membrane, ER nuclear, 26 0 23,6455 0 0,0010227 2 1 2 1 134 Tight junction protein ZO-2 TJP2 cell adhesion membrane Zinc finger CCCH domain-containing protein RNA 27 0 23,0346 0 0,002062 2 1 2 1 112 ZC3H7B nuclear 7B metabolism Dolichyl-diphosphooligosaccharide--protein 28 0 23,0143 0 4,30E-09 1 3 3 1 94 STT3B ERAD ER glycosyltransferase subunit STT3B cytoplasmic, 29 0 23,5728 0 5,52E-06 1 1 2 1 119 Unconventional myosin-Ic MYO1C cytoskeletal nuclear translation, 30 0 23,0214 0 0,00020092 1 0 2 0 233 HEAT repeat-containing protein 1 HEATR1 RNA nuclear metabolism 31 0 24,0015 0 0,00020596 1 1 2 1 110 WD repeat-containing protein 59 WDR59 metabolism Probable ATP-dependent RNA helicase RNA 32 0 23,6688 0 0,00079983 0 2 2 1 90 DDX27 nuclear DDX27 metabolism

- 236 -

Appendix 6: Category V (proteins at least four fold enriched)

14 Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal Heat shock 70 kDa protein cytoplasmic, 1 37,248 38,0163 34,6338 0 808 62 51 41 70 HSPA1A chaperone 1A/1B nuclear cytoplasmic, Heat shock cognate 71 nuclear, 2 34,896 35,9609 32,9485 0,00E+00 365 38 40 30 71 HSPA8 chaperone kDa protein membrane, exosomes Heat shock 70 kDa protein cytoplasmic, 3 30,711 31,8923 29,5806 0,00E+00 127 36 33 25 94 HSPA4 chaperone 4 mitochondrium Large proline-rich protein cytoplasmic, 4 32,719 30,7855 26,4064 0 106 38 28 10 119 BAG6 chaperone BAG6 nuclear

transcription, cytoplasmic, Complement component 1 translation , RNA nuclear, 5 32,36 32,1061 30,1849 2,17E-302 60 10 9 6 31 Q subcomponent-binding C1QBP metabolism, mRNA membrane, protein, mitochondrial metabolism, mitochondrium immunity

BAG family molecular 6 30,872 30,8642 28,4727 1,76E-251 33 12 9 7 24 BAG2 chaperone not annotated chaperone regulator 2 cytoplasmic, nuclear, Heat shock-related 70 kDa 7 27,29 28,8847 26,5464 5,10E-238 16 9 7 7 70 HSPA2 chaperone membrane, protein 2 mitochondrium, exosomes

cytoplasmic, Vacuolar protein sorting- 8 30,833 30,1861 23,2049 3,94E-209 46 21 17 3 92 VPS35 transport membrane, associated protein 35 exosome cytoplasmic, E3 ubiquitin-protein ligase Ubiquitination/SUMO, 9 30,157 30,5403 27,4961 9,47E-189 44 18 15 12 35 STUB1 nuclear, CHIP DNA repair membrane, ER Poly [ADP-ribose] cytoplasmic, 10 30,198 28,3735 0 2,58E-169 43 23 17 1 96 PARP9 DNA repair, immunity polymerase 9 nuclear

- 237 -

Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal Stress-induced- 11 30,199 31,0953 27,9133 1,36E-161 82 29 28 19 63 STIP1 chaperone cytoplasmic phosphoprotein 1 rRNA 2-O- nuclear, 12 30,8 29,0125 28,6645 2,25E-119 38 15 9 9 34 methyltransferase FBL RNA metabolism membrane, fibrillarin exosomes Very long-chain acyl-CoA 13 28,052 29,02 23,5777 1,26E-113 35 16 16 3 70 SLC27A2 metabolism ER synthetase Elongation factor 1-alpha cytoplasmic, 14 25,684 26,503 24,4823 2,16E-99 7 3 4 2 50 EEF1A2 translation 2 nuclear Hsc70-interacting protein 15 29,569 30,7969 26,2014 2,31E-62 29 13 7 6 41 ST13 ST13P5 chaperone cytoplasmic Putative protein FAM10A5 Exosome complex cytoplasmic, 16 26,647 26,6055 24,2582 3,471E-61 9 6 4 3 30 EXOSC3 RNA metabolism component RRP40 nuclear cytoplasmic, 17 28,473 27,2566 25,8036 1,005E-55 15 8 7 5 39 ATPase ASNA1 ASNA1 transport nuclear, ER cytoplasmic, Vacuolar protein sorting- 18 28,492 27,3756 0 2,707E-54 12 7 5 0 21 VPS29 transport membrane, associated protein 29 Golgi Condensin-2 complex nuclear, 19 26,516 25,902 24,0389 9,233E-51 9 8 4 6 169 NCAPD3 cell cycle subunit D3 membrane cytoplasmic, Vacuolar protein sorting- 20 28,02 27,7688 0 3,31E-50 17 10 6 0 39 VPS26B transport membrane, associated protein 26B Golgi Golgi to ER traffic protein 21 27,337 26,1278 0 5,379E-44 15 6 5 1 37 GET4 transport cytoplasmic, ER 4 homolog Transmembrane protein membrane, 22 28,864 30,2853 26,4916 6,88E-43 15 4 6 5 25 TMEM33 not annotated 33 exosomes cytoplasmic, BAG family molecular Ubiquitination/SUMO, nuclear, 23 26,724 27,3248 0 1,046E-41 16 8 7 1 51 BAG5 chaperone regulator 5 chaperone membrane, mitochondrium cytoplasmic, 24 27,096 25,3863 24,2501 6,043E-40 13 7 4 2 28 Myeloid leukemia factor 2 MLF2 not annotated nuclear, membrane

- 238 -

Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal cytoplasmic, 25 28,113 29,9879 27,5981 1,33E-38 25 9 8 6 45 LanC-like protein 1 LANCL1 signalling nuclear, membrane 26 28,451 27,5324 24,5152 2,444E-36 17 8 7 3 18 Ubiquitin-like protein 4A UBL4A transport cytoplasmic Lon protease homolog, 27 25,308 24,4955 22,806 2,222E-34 4 2 3 3 95 LONP1 protease mitochondrium mitochondrial Nucleolar pre-ribosomal- 28 26,045 25,9514 0 7,331E-33 7 7 5 2 254 URB1 mRNA metabolism nuclear associated protein 1 Transcription factor BTF3 29 25,938 26,1511 0 2,713E-27 3 2 2 2 17 BTF3L4 not annotated not annotated homolog 4 Uncharacterized protein 30 26,84 25,2449 24,5453 2,988E-26 6 3 4 4 22 C7orf50 mRNA metabolism not annotated C7orf50 Programmed cell death 31 24,928 23,9051 0 3,978E-26 5 2 2 2 39 PDCD2L cell cycle cytoplasmic protein 2-like

mRNA metabolism, cytoplasmic, 32 26,045 26,9526 0 3,4E-25 9 4 8 1 209 Protein RRP5 homolog PDCD11 RNA metabolism, nuclear translation

33 24,964 24,0154 0 5,035E-21 6 4 2 1 50 AP-2 complex subunit mu AP2M1 transport membrane Multiple myeloma tumor- 34 26,4 24,7335 0 6,688E-21 5 3 3 1 29 MMTAG2 mRNA metabolism nuclear associated protein 2 35 24,652 24,8493 0 6,578E-20 4 3 2 1 43 Protein AAR2 homolog AAR2 not annotated not annotated U5 small nuclear SNRNP40 36 25,503 24,2924 0 1,567E-17 3 2 3 3 39 ribonucleoprotein 40 kDa mRNA metabolism nuclear DKFZp434D199 protein BRCA1-associated ATM nuclear, 37 25,021 23,4432 22,9097 3,28E-17 5 4 2 2 88 BRAT1 DNA repair activator 1 membrane Telomeric repeat-binding cytoplasmic, 38 24,068 23,5491 0 3,291E-17 3 2 2 1 44 factor 2-interacting TERF2IP transcription nuclear protein 1 28S ribosomal protein S25, 39 25,09 25,1571 0 1,679E-16 6 3 2 1 13 MRPS25 translation mitochondrium mitochondrial

40 25,229 24,7373 0 2,013E-15 5 2 2 1 15 Protein LLP homolog LLPH mRNA metabolism not annotated

- 239 -

Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal Serine hydroxymethyltransferase, 41 25,237 24,6357 0 4,211E-15 2 2 2 1 53 SHMT1 metabolism cytoplasmic cytosolic Serine hydroxymethyltransferase Prolactin regulatory transcription, 42 25,15 24,6204 0 2,208E-14 6 2 3 3 45 PREB nuclear, ER element-binding protein transport Nuclear valosin-containing nuclear, 43 25,232 23,8169 0 4,274E-14 3 3 2 0 65 NVL mRNA metabolism protein-like membrane cytoplasmic, Melanoma-associated 44 22,946 23,098 0 1,523E-12 3 2 2 0 86 MAGED1 transcription nuclear, antigen D1 membrane Ubiquitin carboxyl- terminal hydrolase 45 24,057 23,1927 0 1,351E-11 3 3 2 1 105 USP11 Ubiquitination/SUMO cytoplasmic Ubiquitin carboxyl- terminal hydrolase 11 DNA-directed RNA mRNA metabolism, cytoplasmic, 46 24,155 25,1036 0 1,346E-10 3 2 2 1 16 polymerase II subunit POLR2D translation nuclear RPB4 AP-2 complex subunit membrane, 47 26,027 25,7514 0 5,087E-10 5 4 3 2 17 AP2S1 transport, immunity sigma endosomes Nuclear RNA export factor cytoplasmic, 48 25,39 25,6718 0 2,289E-09 5 4 3 3 70 NXF1 mRNA metabolism 1 nuclear

Condensin-2 complex cell cycle, genome nuclear, 49 24,952 23,7913 0 3,883E-09 2 2 2 1 32 NCAPH2 subunit H2 structure membrane

Bromodomain adjacent to cell cycle, genome 50 24,547 23,0577 22,5368 1,417E-08 5 3 2 3 179 zinc finger domain protein BAZ1A nuclear structure 1A

Exosome complex cytoplasmic, 51 25,146 23,7655 0 5,137E-08 3 2 2 1 19 EXOSC1 RNA metabolism component CSL4 nuclear

Serine/threonine-protein cytoplasmic, 52 24,72 24,462 0 6,998E-08 3 2 3 2 92 phosphatase 6 regulatory PPP6R3 nuclear subunit 3 N-alpha-acetyltransferase 53 23,409 24,9448 0 8,263E-08 3 2 2 1 7 NAA38 mRNA metabolism nuclear 38, NatC auxiliary subunit

- 240 -

Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal cytoplasmic, 54 24,586 24,2722 0 8,266E-08 4 3 2 3 43 Protein pelota homolog PELO cell cycle nuclear Cleavage and 55 25,022 24,7909 0 2,644E-07 3 2 3 1 52 polyadenylation specificity CPSF6 mRNA metabolism nuclear factor subunit 6 Nucleolar complex protein 56 25,869 24,9879 0 3,531E-07 3 2 3 1 93 NOC3L mRNA metabolism nuclear 3 homolog Unhealthy ribosome 57 24,371 24,1229 0 5,102E-07 3 3 2 1 171 biogenesis protein 2 URB2 not annotated nuclear homolog Ribonuclease P protein 58 23,769 23,5937 0 2,195E-06 2 2 2 1 15 POP7 RNA metabolism nuclear subunit p20 DNA/RNA-binding protein cytoplasmic, 59 26,171 27,3642 0 2,988E-06 3 3 3 1 45 KIN cell cycle, DNA repair KIN17 nuclear 60 24,747 24,5003 0 5,371E-06 3 3 3 2 41 Surfeit locus protein 6 SURF6 translation nuclear Nucleolar and coiled-body cytoplasmic, 61 24,405 23,9568 0 1,099E-05 4 3 3 2 74 NOLC1 cell cycle phosphoprotein 1 nuclear Putative 60S ribosomal 62 25,165 25,167 0 3,458E-05 3 2 2 1 6 protein L39-like 5 60S RPL39P5 RPL39 translation cytoplasmic ribosomal protein L39 NE pore membrane protein POM 121C NE POM121C mRNA metabolism, 63 22,262 20,4963 0 6,409E-05 2 2 2 1 125 nuclear, ER pore membrane protein POM121 transport POM 121 U4/U6.U5 tri-snRNP- cytoplasmic, 64 22,795 23,1311 0 9,144E-05 5 2 2 2 90 SART1 mRNA metabolism associated protein 1 nuclear, golgi Monocarboxylate 65 25,587 25,0155 0 9,673E-05 5 2 2 1 54 SLC16A1 transport membrane transporter 1 Mediator of RNA nuclear, 66 24,351 23,9725 0 0,0006535 2 2 2 1 168 polymerase II transcription MED1 transcription membrane subunit 1 Vacuolar protein sorting- associated protein 4A VPS4A VPS4B membrane, 67 23,974 23,9065 0 0,0006825 3 2 2 1 49 Vacuolar protein sorting- cell cycle, transport FIGNL1 endosomes associated protein 4B Fidgetin-like protein 1

- 241 -

Appendix 6: Category V (proteins at least four fold enriched)

Intensity unique peptides MS/MS NT3 paren PEP score MW Protein names Gene names functional annotation localisation NT3 Count NT3 NT3NLS control ΔNLS tal

cytoplasmic, 68 23,972 24,1649 0 0,0013653 2 2 2 2 62 Nardilysin NRD1 protease membrane

- 242 -

Appendix 7: Attached materials

15 Appendix 7: Attached materials

One CD has been attached to this copy of my thesis containing the cross-referenced movies in section 6.2.3. One copy of my publication ‘Functional analysis of nuclear localisation signals in VP1-2 homologues from all herpesvirus subfamilies’ (Hennig et al. 2014) has also been attached to the bound copy of my thesis.

- 243 -

Appendix 8: Permissions for reproduction of published materials

16 Appendix 8: Permissions for reproduction of published materials

- 244 -

Appendix 8: Permissions for reproduction of published materials

NATURE PUBLISHING GROUP LICENSE TERMS AND CONDITIONS May 17, 2015

This is a License Agreement between Thomas Hennig ("You") and Nature Publishing Group ("Nature Publishing Group") provided by Copyright Clearance Center ("CCC"). The license consists of your order details, the terms and conditions provided by Nature Publishing Group, and the payment terms and conditions. All payments must be made in full to CCC. For payment instructions, please see information listed at the bottom of this form. License Number 3631610085213 License date May 17, 2015 Licensed content publisher Nature Publishing Group Licensed content publication Nature Reviews Molecular Cell Biology Licensed content title Orchestrating NE disassembly and reassembly during mitosis

Licensed content author Stephan Guttinger, Eva Laurell and Ulrike Kutay Licensed content date Mar 1, 2009 Volume number 10 Issue number 3 Type of Use reuse in a dissertation / thesis Requestor type academic/educational Format print and electronic Portion figures/tables/illustrations Number of 1 figures/tables/illustrations High•res required no Figures Figure 1 Author of this NPG article no Your reference number None Title of your thesis / Function and transport of a HV encoded Ubiquitin•specific protease in dissertation virus entry and assembly Expected completion date May 2015 Estimated size (number of 240 pages) Total 0.00 EUR Terms and Conditions

- 245 -