Adenoviral Small Non-Coding Rnas

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1216

Adenoviral small non-coding RNAs

A Structural and Functional Charaterization

WAEL KAMEL

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6206 ISBN 978-91-554-9556-5 UPPSALA urn:nbn:se:uu:diva-283080 2016 Dissertation presented at Uppsala University to be publicly examined in B/A1:111a, BMC, Uppsala, Thursday, 2 June 2016 at 13:15 for the degree of Doctor of Philosophy (Faculty of Medicine). The examination will be conducted in English. Faculty examiner: Prof. Ben Berkhout (Academisch Medisch Centrum University of Amsterdam, Netherlands).

Abstract Kamel, W. 2016. Adenoviral small non-coding RNAs. A Structural and Functional Charaterization. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1216. 49 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9556-5.

Since their discovery in 1953, adenoviruses have significantly contributed to the understanding of virus-host cell interactions, including mechanistic details of cellular processes such as cell cycle control and alternative RNA splicing. Among the first characterized adenoviral genes were the virus-associated RNAs (VA RNAI/II), which are produced in massive amount during a lytic infection. The VA RNAs perform multiple functions and are required for a successful adenovirus life cycle. More recently, it was shown that the VA RNAs are processed into small viral miRNAs, so-called mivaRNAs, which interfere with the function of the cellular RNAi/ miRNA machinery. In papers I and II, we focused on a structural and functional characterization of the mivaRNAs using two approaches. Firstly, we created a model system where the predicted miRNA- like function of mivaRNAI could be investigated, without interfering with other VA RNA functions. This was accomplished by construction of recombinant adenoviruses, in which the seed sequence of mivaRNAI was altered. The results showed that in cell culture experiments the mivaRNAI seed sequence mutants grew as the wild type virus, suggesting that the mivaRNAs are not required during the lytic phase of an adenovirus infection. Secondly, we showed that the VA RNAs from different human adenoviruses (Ad4, Ad5, Ad11 and Ad37) undergo the same type of Dicer-dependent processing into mivaRNAs, which subsequently are loaded onto the RNA induced silencing complex (RISC), albeit with different efficiencies. In paper III, we demonstrated that the promoter proximal region of the adenovirus major late promoter (MLP) produces a novel non-canonical class of small RNAs, which we termed the MLP-TSS-sRNAs. Surprisingly the MLP-TSS-sRNA maintains the m7G-cap structure while bound to Ago2 containing RISC. These complexes are functional suppressing expression of target mRNAs with complementary binding site. Most importantly, the MLP-TSS-sRNA limits the efficiency of viral DNA replication probably through a targeting of the E2B mRNAs, which are transcribed in the antisense orientation. In conclusion, the MLP-TSS-sRNA represents the first viral small RNA, which has been shown to have a function as a regulator of an adenovirus infection.

Keywords: non-coding RNA miRNA mivaRNA canonical Dicer Ago2 TSS m7G Cap Adenovirus replication E2B

Wael Kamel, Department of Medical Biochemistry and Microbiology, Box 582, Uppsala University, SE-75123 Uppsala, Sweden.

ISSN 1651-6206 ISBN 978-91-554-9556-5 urn:nbn:se:uu:diva-283080 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-283080)

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Kamel W, Segerman B, Oberg D, Punga T, Akusjarvi G (2013) The adenovirus VARNA-derived miRNAs are not essential for lytic virus growth in tissue culture cells. Nucleic Acids Res 41: 4802-4812.

II Kamel W, Segerman B, Punga T, Akusjarvi G (2014) Small RNA Se- quence Analysis of Adenovirus VA RNA-Derived MiRNAs Reveals an Unexpected Serotype-Specific Difference in Structure and Abundance. PLoS One 9: e105746.

III Kamel W, Akusjarvi G (2016) Characterization of a non-canonical 5’ capped adenoviral small RNA that is functionally active. Manuscript.

Reprints were made with permission from the respective publishers.

Other Papers not included in thesis

I. Punga T, Kamel W, Akusjarvi G (2013) Old and new functions for the adenovirus virus-associated RNAs. Future Virology 8: 343-356.

II. Inturi R, Kamel W, Akusjarvi G, Punga T (2015) Complementation of the human adenovirus type 5 VA RNAI defect by the Vaccinia virus E3L protein and serotype-specific VA RNAIs. Virology 485: 25-35.

III. Pickl JM, Kamel W, Ciftci S, Punga T, Akusjarvi G (2015) Opposite expression of CYP51A1 and its natural antisense transcript AluCYP51A1 in adenovirus type 37 infected retinal pigmented epithelial cells. FEBS Lett 589: 1383-1388.

Members of the committee

Opponent Ben Berkhout, Professor Academisch Medisch Centrum University of Amsterdam, Netherlands

Members of the committee Göran Magnusson, professor Department of Medical biochemistry and Microbiology Uppsala University

Gerhart Wagner, professor Department of Cell and Molecular biology Uppsala University

Marie Öhman, professor Department of Molecular Biosciences Stockholm University

Table of Contents

Introduction ...... 11 Noncoding RNA: the rise of the underdog ...... 11 MicroRNA pathway ...... 13 Canonical miRNA production ...... 14 Non-canonical miRNA production ...... 14 MiRNA target recognition and mechanism of action ...... 16 Viruses and host gene regulation: lessons from the enemy ...... 17 Virus interplay with the cellular miRNA machinery ...... 19 Adenovirus ...... 23 Virus-associated RNA (VA RNA): a multifunctional non-coding RNA .... 26 Background ...... 26 A- VA RNA mediated-suppression of the innate immune response .... 28 B- VA RNA mediated suppression of the miRNA pathway ...... 30 Present investigation ...... 31 Paper I: The adenovirus VA RNAI-derived miRNAs are not essential for lytic virus growth in tissue culture cells ...... 31 Paper II: Small RNA sequence analysis of adenovirus VA RNA-derived miRNAs reveals an unexpected serotype-specific difference in structure and abundance ...... 33 Paper III: Characterization of a non-canonical 5’ capped adenoviral small RNA that is functionally active ...... 36 Conclusions and future perspectives ...... 38 Acknowledgment ...... 40 References ...... 43

Abbreviations

NcRNA Non-coding RNA MiRNA MicroRNA Pre-miRNA Precursor miRNA Pri-miRNA Primary miRNA 3’UTR 3’ untranslated region Xpo5 Exportin5 HAd Human adenovirus Ad Adenovirus Pol II RNA polymerase II Pol III RNA polymerase III VA RNA Viral Associated RNA nts Nucleotides Ago Argonaute PKR Protein kinase R OAS 2′–5′-Oligoadenylate Synthetases PTP Preterminal protein Adpol Adenovirus DNA polyermase siRNA Small interfering RNA C19MC The chromosome 19 miRNA cluster RISC RNA induced silencing complex miRISC micoRNA induced silencing complex 5p 5’strand 3p 3’strand HSV Herpes Simplex Virus HSUR HSV U-rich RNA TRBP HIV-1 TAR RNA binding protein MHC Major histocompatibility complex PHAX Phosphorylated adapter RNA export protein MLP Major late promoter TSS Transcription start site DGCR8 DiGeorge syndrome chromosomal region 8

Introduction

My PhD thesis is a humble attempt to characterize the interactions between a mammalian virus and the microRNA pathways. Therefore, I believe that in order to properly present my work, I will divide the introduction into three main parts; first, a description of noncoding RNAs with focus on the mi- croRNA pathway. Followed by description of the impact of virology on RNA research. Then finally, I will briefly describe the adenovirus model system and its previously characterized interactions with the microRNA pathway.

Noncoding RNA: the rise of the underdog Noncoding RNA (ncRNA) represents a wide spectrum of different RNA molecules with one main common feature, which is the lack of a protein- coding capacity or a function that is independent of the potential protein- coding capacity (Figure 1). This RNA category consists of several RNA classes that includes transfer RNA (tRNA), ribosomal RNA (rRNA), mi- croRNA (miRNAs), small interfering RNAs (siRNAs), P-element-induced wimpy testis (PIWI)-interacting RNAs (piRNAs), small nucleolar RNAs (snoRNAs), small nuclear RNAs (snRNAs), and long ncRNAs (LncRNA). The rise of ncRNA as an essential element in gene regulation reflects a strong shift in our understanding of the central dogma, in which the RNA was proposed to merely be a “messenger” between DNA and a protein. Although Jacob and Monod proposed the involvement of ncRNA in regulation of gene expression in 1961 [1], ncRNA (with exception of the classical tRNA and rRNA) received initially little attention from the scientific community. This negligence was partially due to the lack of high throughput technologies to detect the low abundant classes of ncRNA, which are often considered transcriptional noise rather than meaningful transcripts. The initial wave of revolution in the ncRNA field started with the discovery of the snRNAs [2], which are part of ribonucleoprotein complexes (snRNPs) that act as essential factors in RNA splicing [3]. Later on, a novel group of small RNAs that are strictly localized in the nucleolus was characterized, the so- called snoRNAs. The snoRNAs guide chemical modifications of other RNA molecules such as tRNAs, rRNAs and snRNAs, which is important for their function [4,5]. During the early 80s, several groups, including Cech’s and

11 Altman’s, elegantly demonstrated that RNA is able to catalyze enzymatic reactions, a feature that was believed to be restricted to proteins. This group of RNA was later known as ribozymes. Ribozymes are involved in the catalysis of several RNA processing reactions, such as RNA splicing (the self- splicing introns) and the processing of tRNA (the removal the leader sequence from pre-tRNA by RNase P) [6,7]. The discovery of ribozymes was indeed a landmark in the development of the ncRNA field, since it did not just directly challenge the central dogma but also opened the way for other ncRNA species to be discovered. The pioneering work on ribozymes was acknowledged by the scientific community in 1989 when Sidney Altman and Thomas R. Cech shared the Noble Prize in chemistry[8]. In the early 2000s with the development of high throughput technologies, there was a peak of discoveries of different classes of novel ncRNAs, such as miRNA, siRNA, piRNA and large heterogeneous group of LncRNA. The- se findings were sparked by the completion of human genome project, because it became evident that the protein coding capacity of the human genome is minor (about 2%), whereas the rest of the genome has probably a non-protein coding function [9-11]. Additionally, one of the interesting out- comes of the ENCODE project, which aimed to functionally characterize the human genome, was that the large proportion of the human genome is ac- tively transcribed, and the major output of this transcription is ncRNA [12]. Further, it has been shown that there is a positive relation between organism complexity and the proportion of ncRNA in the genome [13]. Altogether, this suggests that different ncRNA species introduce multilayers of gene regulation, which play a crucial role in defining the developmentally complex organisms such as humans.

12 mRNA

Coding RNA

RNA

Non-coding RNA

> 200 nt < 200 nt

Long ncRNA Small ncRNA snoRNA rRNA

tRNA snRNA LincRNA other

Intronic piRNA LncRNA others

Antisense siRNA LncRNA miRNA

Figure 1. Classification of different RNA species into coding and non-coding RNA

MicroRNA pathway MiRNAs form a class of small RNAs, around ~ 22 nucleotides (nt) long, that were initially discovered from forward genetic screens of developmental timing mutant in the nematode C. elegans [14]. A few years after the initial discovery, several reports demonstrated that miRNAs constitutes a large group of genes that are conserved among insects, nematodes and humans. Later on, miRNAs have been detected in a wide variety of organisms including brown algea, mollusk, tunicates, sea lamprey, vertebrates, insects, amoe-

13 ba, monocots, dicots, large DNA viruses, and recently in T. gondii, a unicel- lular protozoan parasite [15-19]. According to the miRNA database, www.mirbase.com, around 2500 miRNAs had been identified in humans [20]. The majority of the human miRNAs are encoded within the intronic part of host genes produced by RNA polymerase II transcription [21,22]. However, some miRNA genes are produced by RNA polymerase III transcription. For instance C19MC, the largest known cluster of miRNAs, that contains 46 tandemly repeated miR- NA genes are transcribed by RNA polymerase III [23].

Canonical miRNA production MiRNAs are transcribed in the nucleus as long transcripts, called primary miRNA (pri-miRNA), with several stem loops flanking single strand RNA sequences. The pri-miRNA will undergo a two-step sequential processing, which involves two RNAse III type endonucleases, Drosha and Dicer [24,25]. The first processing step takes place in the nucleus by the micropro- cessor complex containing Drosha with its cofactor DGCR8 (in mammals). During this step the pri-miRNA is processed into an approximately 70nts long dsRNA with imperfect stem loop structures, the so-called precursor miRNA (pre-miRNA). The pre-miRNA is exported to the cytoplasm using Exportin 5, a RanGTP-dependent dsRNA-binding nuclear export factor [26]. Once in the cytoplasm, the pre-miRNA undergoes a second processing step, where the Dicer enzyme, together with its cofactor the TAR RNA binding protein (TRBP), cleaves near the terminal loop in the pre-miRNA generating a 22 base pair miRNA duplex. Finally the miRNA duplex will be allocated to a guide strand selection process. During this process the miRNA strand with the lowest thermodynamic stability at the 5’ end will be selected as the guide strand and loaded onto the miRNA induced silencing complex (miR- ISC), whose core components are the argonaute family of proteins (AGO1- 4). The other strand (the passenger strand, miRNA*) will in most cases probably be degraded [27].

Non-canonical miRNA production In 2007, soon after the establishment of the canonical miRNA-processing pathway, a new group of miRNAs, originates from protein coding genes, was shown to bypass Drosha cleavage in flies, nematodes [28,29] and later on in mammals [30]. These miRNAs are processed from short introns that are spliced and debranched into pre-miRNA like structures, so called mirtrons (Figure 2). Mirtrons then rejoin the canonical miRNA pathway for export by Exportin 5 and Dicer cleavage. A second group of Drosha- independent miRNAs originates from the transcription start site of protein coding genes, so-called TSS-miRNAs [31]. The pausing of RNA polymerase

14 II defines the 3’ end of these miRNA generating a pre-miRNA with a 5’capped tail. The cap structure has been suggested to be removed by an decapping/exonuclease activity resulting in the formation of hairpin like structured RNAs that can join the canonical miRNA pathway for export, Dicer processing and RISC assembly (Figure 2). Another example is pre- miR-484, which is produced as a 5’ capped hairpin RNA. Unlike the TSS- miRNAs, the 5’end cap is not cleaved off instead it binds to the cap-binding complex and is exported to the cytoplasm either through PHAX facilitated export or through Exportin1 (it can also be exported by Exportin 5) [32]. Once in the cytoplasm, Dicer will process pre-miR-484 into a miRNA duplex where the uncapped 3’strand will be preferentially incorporated onto RISC (Figure 2). So far miR-451 is the only example of a Dicer independent cellular miR- NA. Like most cellular pre-miRNAs, pre-miR-451 is generated by Drosha cleavage, which produces an exceptionally short hairpin (17 base-pair long). Pre-miR-451 is exported by the Exportin5 pathway but bypasses Dicer cleavage and binds directly to Ago2 [33,34]. Ago2 (the only mammalian Ago with a slicer activity) will cleavage pre-miR-451 generating a long guide strand (around 30nts, 5p-miR-451) and shorter passenger strand (3p- miR-451). The 5p-miR-451 is finally trimmed by the poly (A)-specific ribonuclease (PARN) into a 23nts long miRNA (Figure 2) [35].

15 Cellular miRNA biogenesis

Nucleus Canonical pathway Cytoplasm

Exportin 5 Argonaute binding (A)n Dicer

(A)n pre-miRNA Cleavage pri-miRNA Drosha/DGCR8 cleavage

Drosha-independent non-canonical pathway Exportin 5 1) intron (mirtron) debranching Lariat Merge with formation the canonical pathway exon exon Spliceosome

Exportin 5

2) Argonaute Pol II transcription Dicer binding and termination? Exportin1 m7G-cap Cleavage PHAX

CBC

3) ? Pol II transcription m7G-cap Merge with ? the canonical pathway and per-termination ? Dicer-independent non-canonical pathway

(A) n PARN Drosha/DGCR8 pri-miR 451 Ago2 binding trimming cleavage and slicing

Figure 2. Different routes for cellular miRNA biogenesis in animals, adapted from[36]

MiRNA target recognition and mechanism of action After the discovery of the miRNAs and realization of their abundance in diverse living organisms, questions were raised about their mode of action and function. Initial clues about the action of miRNAs originated from the observation that the lin-4 RNA, the first characterized miRNA, had partial

16 complementary with multiple conserved sites in the 3’ untranslated region (3’UTR) of the lin-14 mRNA [14,37]. Subsequent studies concluded that miRNA functions as a linker molecule guiding the miRISC components to their target mRNAs. The miRNA-target recognition is dependent on Wat- son–Crick base pairing between the 5' region of the miRNA, centred on nucleotides 2–7, the so called seed sequence, and the 3’ UTR of the target mRNA [38]. However, recently several exceptions to this rule have been identified experimentally. For example the let-7 miRNA has two targeting sites in the lin-41 mRNA, and neither one has a perfect complementary to the seed sequence of let-7 [39]. In addition to the seed sequence base pairing, central mismatches or bulges often occur with some compensatory base pairing between the 3’ end of the miRNA and the mRNA. However it was shown that this supplementary base pairing are less conserved and not crucial to miRNA function [40]. Soon after the miRNA-mRNA interaction is established in mammalian cells, the miRISC downregulates the expression of the target mRNAs by two different mechanisms. They either direct an endonucleolytic cleavage or through translation repression of the mRNA. The direct cleavage of an mRNA only takes place when there is a perfect complementary between the miRNA and its target and it is mediated by Ago2, the only member of the mammalian Ago family with endonucleolytic activity. As a consequence of the flexibility in target recognition, one miRNA can bind to hundreds of mRNAs, and an mRNA can be targeted by multiple miRNAs. It has been estimated that the miRNA pathway regulates the expression of around 30-60% of human genes, and as a consequence, it is involved in diverse cellular and disease processes such as proliferation, mor- phogenesis, apoptosis and differentiation [41].

Viruses and host gene regulation: lessons from the enemy I believe the motivation behind virus research can be summarized in two main goals: either eradicate all viruses from the surface of planet earth or utilize viruses as exploratory tools to learn more about their hosts, which in some occasions is us, humans. The latter motivation seems more appealing, given the unique properties of viruses, which I will try to summarize in the next few points. Viruses can infect almost all life forms in which they inter- act with all cellular processes. This makes viruses powerful tools that can be utilized in wide range of model organisms and research fields. The diversity of the genetic material of viruses that are capable of infecting the same host (Table 1), provides different perspectives for the interaction with the host cells biosynthetic machineries. Virus genomes are in general small, thus

17 making them easy to manipulate and edit. Finally, viruses have a relatively short life cycle in tissue culture cells, during which they produce a limited number of transcripts in high abundance that can be easily isolated and characterized. This feature was especially highly appreciated during earlier times when sensitive and high-throughput techniques to characterize cellular transcripts were not available.

Table.1 The Baltimore classification [42] Group Genome type Example I dsDNA Adenovirus II ssDNA Adeno-associated virus III dsRNA Retrovirus IV ssRNA, positive sense Poliovirus V ssRNA, negative sense Influenza virus VI dsRNA, DNA intermediate HIV VII dsDNA, RNA intermediate Hepatitis B virus

Taking altogether, it is not surprising that several key breakthroughs in our understanding of mammalian gene regulation originated from studies using viruses as model systems (Figure 3). For instance, capping of cellular mRNAs and the importance of the cap for efficient translation and stability originated from in vitro studies of reoviurs mRNA translation [43-46]. The notion “one mRNA encodes one protein” was proven wrong in animals through a series of milestone discoveries starting with the discovery of split genes and mRNA “alternative splicing” [47-49] and “alternative poly (A) sites usage” in the processing of the adenovirus major late transcript. Fur- ther, leaky scanning of simian virus (SV40) mRNA, [50] and Ribosomal frame shifting in RNA viruses, described additional mechanisms how multiple proteins could be produced from a single mRNA. There are many more examples demonstrating how viruses have been indispensable tools for understanding the molecular events of their host.

18 1975 !"##$%&'()'*+,-' Reovirus and Vaccinia virus

1977 -./01%"230'4#.$5$%&' Adenovirus

1985 61"%4."2(%".')1"*047$8$%&'$%'09:"1;(/04'' Retroviruses

1988 <%/01%".'+$=(4(*0'>%/1;'?$/0' Picornaviruses

1990s +,-$' Antiviral immune responses

1997 ,95.0"1'*+,-'0@#(1/' Retroviruses

Figure 3. Timeline representing key post-transcriptional modifications that were first defined using viral systems. (*) pinpoints discoveries lead to Nobel Prize.

Virus interplay with the cellular miRNA machinery Bearing in mind the involvement of the miRNA pathway in many cellular processes, it is not surprising that viruses has evolved a variety of strategies to manipulate the miRNA pathway to its own advantage. For example, hijacking of the miRNA machinery to produce viral miRNA, suppressing the production and promoting the degradation of cellular miRNA, and other mechanisms that will be discussed in detail in the following section.

Viral miRNAs Three year after the first large scale discoveries of more than hundred miR- NAs in D. melanogaster and C. elegans [51-53], Pfeffer et al. reported the production of a number of viral miRNAs in EBV (Epstein-Barr virus) latently infected B cells [54]. Later on, several reports showed that other DNA viruses also encoded for their own miRNAs [55-58]. Most of the known viral miRNAs are produced following the canonical miRNA pathway. For example, EBV encodes for 25 pre-miRNAs and 44 mature miRNAs originat- ing from two main clusters (BART and BHRF1), KSHV (Kaposi’s sarcoma- associated herpesvirus) encodes 12 pre-miRNAs from the latent region generating around 22 mature miRNAs [19]. However viruses can also be crea- tive in their ways of producing miRNA by non-canonical Drosha- independent mechanisms as shown in Figure 4. For instance in murine "- herpesvirus 68 (MHV68), the viral pre-miRNA is transcribed by RNA polymerase III as a read-through product of a tRNA gene. This tRNA-pre- miRNA is cleaved by the host tRNAse Z enzyme, bypassing Drosha cleavage and generating a pre-miRNA that join the canonical miRNA export and

19 processing pathway [59]. Similarly, in Herpesvirus saimiri (HVS) viral miRNAs are encoded as read-through extension of viral snRNA like genes. The host integrator complex processes the viral snRNA-pre-miRNA releas- ing the pre-miRNA [60]. Other viruses, such as adenoviruses and bovine leukaemia virus (BLV), encode for genes that are transcribed by RNA polymerase III forming short hairpin RNA (shRNA) that resemble a pre-miRNA, thereby bypassing Drosha cleavage but requiring Exportin5 and Dicer processing [61].

20 Viral miRNA biogenesis

canonical pathway Nucleus Cytoplasm

Exportin 5 Argonaute (A)n Dicer binding Drosha/DGCR8 (A)n cleavage Cleavage pri-miRNA pre-miRNA

Drosha-independent non canonical pathway

Integrator complex HVS miRNA

Other nucleases SM Viral snRNA-moiety

MHV68 miRNA tRNase Z

Other nucleases

Merge with the canonical pathway

Viral tRNA-moiety

BLV and Ad miRNA

Pol III transcription and termination

Viral shRNA Figure 4. Different routes for miRNA biogenesis in mammalian virueses, adapted from [62].

21 Cellular miRNA: the foe and the friend

The foe RNA interference (RNAi), which was initially discovered in C. elegans, is a mechanism similar to the miRNA pathway that leads to RNA degradation in a sequence specific pattern. RNAi is considered as an important antiviral innate immune response, particularly in plants, since it is induced upon virus infection that results in the production of virus-derived double stranded RNA (dsRNA). The dsRNA is processed by Dicer, similar to the pre-miRNA, into small dsRNAs of ~21–25 bp in length, so called short interfering RNA duplex (siRNA duplex). The guide strand of the siRNA duplex is loaded onto the RNA induced silencing complex (RISC) and will direct the cleavage of complementary mRNA targets [63]. Both miRNAs and siRNAs share properties in the way they are produced and their mode of action, and the only clear distinction between them is the origin of the dsRNA [64]. Since the discovery of RNAi, it has remained highly controversial what significance this mechanism plays in mammals. Although RNAi can be arti- ficially induced in mammalian cells, the evidence demonstrating that it is part of the normal antiviral innate immune response has been weak. Howev- er, a recent report demonstrated that in Nodamura virus infected mouse cells, viral small RNAs accumulated in a Dicer-dependent manner and bound to RISC. Further, accumulation of these viral small RNAs significantly reduced Nodamura virus growth [65]. On the other hand, as mentioned above, cellular miRNAs represents an expanding heterogeneous family of small RNAs with a wide range of targets, mainly due to the low requirements for target recognition. Also there are more than 1000 miRNAs that show tissue specific expression [66]. This means that it is highly likely that cellular miRNAs can target mammalian viruses, especially for viruses infecting several cell types. Thus, it is not surprising that many viruses have evolved several strategies to suppress the miRNA machinery. The viral-induced suppression of the miRNA pathway can be global or specific to certain miRNAs. For instance in poxviruses, the viral poly (A) polymerase (VP55) induces a specific degradation of cellular miRNAs, with no sequence bias [67]. Similarly, a variety of mammalian viruses encode for global RNAi/microRNA silencing suppressors (RSS), reviewed in [68]. On the other hand HVS encodes for ncRNAs that resembles the Sm-class of snRNAs, the so-called HSV U-rich RNAs (HSURs). These RNAs accumulate at high levels in latently HSV infected T cells [69,70]. The HSURs can bind to the Sm family of proteins and form complexes similar to the snRNPs [71]. However, the HSURs are not involved in RNA splicing. In- stead, the HSURs acts as sponge RNAs, sequestering a subset of cellular miRNAs through base pairing, thereby causing miRNA degradation [72]. It

22 has been proposed that the virus has reprogramed the host splicing machinery for miRNA degradation [62].

The Friend Viruses not always consider cellular miRNA as bad news that they need to get rid of. There are several examples in which viruses elegantly utilize the host miRNA machinery for their own benefit, as is exemplified by Hepatitis C virus (HCV) and miR-122. HCV is a RNA virus with positive sense single stranded genome, which replicates mainly in the liver. MiR-122 is a classical example of tissue specific miRNA, highly upregulated in liver (constitutes around 70% of the total miRNA pool)[73]. In 2005, Jopling et al. showed that miR-122 was required for an HCV infection since depletion of miR-122 from liver cells resulted in significant decrease of HCV growth [74]. Moreo- ver they identified two binding site for miR-122 in the 5’ region of HCV genome, which appear to be crucial for virus growth. The exact mechanism on how binding of miR-122 to the 5’UTR of HCV contributes to viral growth is still unknown. One possibility that was suggested by Li et al. is that miR-122 binding act as cap protecting the HCV genome from 5’-3’ exonuclease activity [75]. HCV and miR-122 represents a highly unique case for two main reasons; firstly this is the only example, as far as I know, where the tropism of a mammalian virus is determined by a cellular miRNA. Secondly, it represents a situation in which the classical function of a miRNA is reversed. Thus, instead of promoting degradation and instability of a target RNA, or suppressing translation, it functions a positive factor protecting and stabilizing HCV. This represents a second example (the first example was using the host cell splicing machinery for miRNA degradation, mention in the above section) showing us how clever viruses can be in hijacking a host cell machinery and reprogram it for a different purpose.

Adenovirus Adenoviruses are common in nature, infecting a wide range of living organisms including mammals (Mastadenovirus), birds (Aviadenovirus), reptiles (Atadenovirus), amphibians (Siadenovirus) and fish (Ichtadenovirus) [76]. Adenoviruses are double stranded DNA viruses, named after adenoid cells where they were first isolated in 1953 during the search for the ‘virus of the common cold’ [77]. An adenovirus is characterized by its non-enveloped icosahedral capsid of 90 nm in diameter consisting of 240 hexons and 12 pentons, with a protruding fiber from each penton base [78]. Today there are more than 60 human adenovirus genotypes characterized [79], which are classified into seven species (A through G) mainly based on serology and genome sequences. Adenoviruses mainly target terminally

23 differentiated epithelial cells in various organs. However, different serotypes show different tissue tropism: species A and F mainly infect the gastrointes- tinal tract, species B, C and E infect the respiratory tract and urinary tract while species D viruses are the major cause of ocular infections [80]. Forty years after its discovery, the first adenovirus receptor, the cox- sackievirus adenovirus receptor (CAR), was characterized [81]. It was shown that most human adenovirus serotypes, except for serotypes from species B, bind to the extracellular domain D1 in CAR through the fiber knob. Moreover, the penton-base of all adenovirus species, except members of species F, possesses an arginine-glycine-aspartate (RGD) motif that binds to cellular integrins acting as a co-receptor promoting viral entry into the cells. Additionally it has been reported that the membrane cofactor CD46 is required for binding of Ad3, 11, 16, 21, 35, 50 (species B1 and B2), as well as Ad37 (species D) to epithelial cells [82,83]. After attachment of adenovirus to the cell surface, the virus together with its receptor assembles into clathrin-coated pits, which are internalized forming endosomes. Adenoviruses can be observed in the endosomes already 5 min after attachment to its cell surface receptor [84,85]. The virus undergoes a partial uncoating inside the endosome, where capsid proteins are removed via proteolytic degradation and selective dissociation. Adenovirus escape form endosome is mediated by viral protein VI, which exhibits pH- independent membrane lytic activity [86].

Figure 5. Map of adenovirus genome and transcriptional units. It consists of the five early transcriptional units (E1A, E1B, E2, E3, and E4), and the major late transcription unit [MLP] (L1–L5)). Kindly provided by G. Akusjärvi.

24 Finally adenovirus finds its way to the nucleus by moving on the intracellu- lar network of microtubules. Once the partly dismantled adenovirus capsid arrives at a nuclear pore the viral genome is passed into the nuclear envi- ronment. The adenoviral genome is about 30-38 kbp in length and encodes for some 40-50 proteins. The genome is divided into several transcriptional units according to their time of expression during the viral life cycle. There are five early (E1A, E1B, E2, E3 and E4), two delayed early (pIX and IVa2) and one late transcriptional unit, the so-called major late transcription unit (MLTU). Transcription of all adenoviral genes is mediated by RNA polymerase II, except for the VA RNA genes, which are transcribed by RNA polymerase III [87] (Fig. 5). Proteins encoded by the adenoviral E1A unit are the first to be expressed and can be detected as early as 1 hours post infection. Due to alternative splicing, the E1A gene produce 5 alternatively spliced transcripts; 13S, 12S, 11S, 10S, 9S [88]. Most of the characterized E1A functions are carried out by the proteins encoded by the 13S and 12S mRNAs (the 243R and 289R proteins), while the products of the other minor transcripts have no clear function. The E1A proteins functions as transcription factors activating the other early viral transcription units, E1B, E2, E3, E4 and E1A itself [89]. Also the E1A proteins induce G1 to S phase entry by binding and suppressing the function of the retinoblastoma (RB) protein. This is vital for the virus to be able to initiate viral DNA replication [90,91]. The E1B proteins are vital for virus replication by blocking virus-induced apoptosis. The E1B-55K protein binds and blocks p53 activation [92]. While the E1B-19K protein sequesters the activated proapoptotic proteins Bcl-2–associated X protein (BAX) and Bcl-2 homologous antagonist-killer protein preventing pore formation in the mitochondria and the release of more proapoptotic proteins [93]. The E2 region encodes for the viral proteins required for viral DNA replication, which includes the preterminal protein (pTP), which acts as a primer for initiation DNA replication, the adenovirus DNA polymerase (Ad- pol) and the 72K single-stranded DNA binding protein [94]. In order to ensure the success and the completion of the viral infection, the proteins encoded from the E3 region act to counteract the host antiviral defence mechanisms. The E3-gp19K protein binds newly synthesized MHC class I proteins in the endoplasmic reticulum, thereby preventing their translocation to the cell surface. As a result, E3-gp19K prevents viral antigen presentation through the MHC class I pathway which would otherwise cause the recognition and disruption of the infected cells by cytotoxic T-cells [95]. The E3- 10.4K and 14.5K proteins form the E3 receptor internalization and degradation complex (RID). E3 RID is an integral membrane protein complex capable of suppressing apoptosis induced by TNF-α (tumor necrosis factor alpha), FasL (Fas ligand) and TRAIL (tumor necrosis-related and apoptosis- induced ligand) [96]. This suppression is mediated by clearance of the

25 chemokine receptors from the cell surface and promoting their degradation in the lysosome. Finally E3 RID together with E3 14.7K inhibits TNF- mediated release of inflammatory mediator arachidonic acid through blocking translocation phospholipase A2 to the cell membrane. The E4 region is predicted to encode for seven polypeptides through a complex set of mRNAs generated by alternative splicing. These proteins are involved in viral DNA replication, viral mRNA splicing and transport, shut-off of cellular protein synthesis and regulation of apoptosis [87]. Activation of viral DNA replication is the demarcation point separating the early from the late phase of the viral life cycle. The MLTU produces a primary transcript of around 30.000 nucleotides. This primary transcript has five different polyadenylation sites generating five different families of late mRNAs (L1 to L5, Fig. 5), which then through alternatively splicing generates at least 18 different cytoplasmic mRNAs [97]. All MLTU mRNAs are tagged with a 203 nucleotides leader sequence at their 5’ end, the so-called tripartite leader. The presence of the tripartite leader enable the late adenoviral mRNAs to be efficiently translated independently of an active cap binding complex (eIF4F) [98]. Since late during infection the adenovirus L4- 100K protein displaces the eIF4E kinase Mnk1 from eIF4F, all mRNAs that relies on cap-dependent translation will not be translated. This will affect translation of both cellular mRNAs and adenoviral mRNAs lacking the tripartite sequences [99]. Finally, the newly synthesized adenoviral proteins are imported back to the nucleus where the virion assembly takes place. In tissue culture cells the whole infection cycle takes 24-48 hours and ends with the lysis of the infected cell and release of 104-105 new viral particles per cell.

Virus-associated RNA (VA RNA): a multifunctional non-coding RNA Background In 1966, Reich et al. reported that enormous amounts of a viral small RNA accumulated in Ad2 infected cells, which was designated as the virus- associated RNA (VA RNA) [100]. Ten years later, it was demonstrated that Ad2 in fact encodes for two VA RNA species. VA RNAI, the major species that was previously identified by Reich et al., and VA RNAII, the minor species that accumulates at lower copy number during an infection [101,102]. Both VA RNAI and II are RNA polymerase III transcripts, around 160 nucleotides long. The VA RNA genes are regulated by intragenic control elements, similar to tRNA genes, which are know as box A and box B elements [103,104]. The location and function of these elements have been thoroughly mapped by mutational analysis [105-107]. It was proposed from the mutational studies that the B box is more crucial for efficient VA

26 RNAI transcription, since it was less tolerable to mutations compared to the A box [106,107]. This is due to the importance of box B as a binding site for the cellular transcription factor TFIIIC, which is required to recruit RNA polymerase III to the VA RNA transcription unit [108,109]. During the early phase of infection, both VA RNAI and II are transcribed at similar rate. However with the progression of the infection and the start of viral DNA replication, the amount of viral DNA templates available for transcription increases. Due to the competition for RNA polymerase III transcription factors the VA RNAI promoter, which is intrinsically stronger than the VA RNAII promoter, will be preferentially transcribed and accumulates to 20-40 fold higher levels compared to VA RNAII late during infection [102]. VA RNAI transcription has two initiation sites, an A and a G start which are three nucleotides apart [110], VA RNAI (G) represents about 75% of the total VA RNAI population. Adenoviruses that lack two base pair located 23- 25 upstream of VAI (A) start site only accumulate the VA RNAI (G) species with no detectable change on virus growth in tissue cultures [111]. All identified adenoviruses, except for some avian adenoviruses, have at least one VA RNA gene. The large majority (more that 80%) of human adenovirus serotypes have two VA RNA genes. The nucleotide sequence of the VA RNAs varies significantly among different adenovirus species, while it is highly homologous between serotypes within the same species [112]. It was reported that there are around fourteen short homologous regions in the VA RNA sequence across different species. The longest homologous region corresponds to the internal transcription signals, the A box and the B box. Most of the other homologous regions are involved in maintain the RNA secondary structure. Surprisingly, the secondary structure of the VA RNAs is well conserved between different species. The VA RNA secondary structure can be divided into three major subdomains: the terminal stem, the central domain and the apical stem (Figure 6). This structure is maintained by G-C rich, mutually complementary sequences at the 3’ and 5’ ends of the VA RNAs, in addition to a highly conserved based-paired tetranucleotides sequence, GGGU/ACCC, that is located in the central domain[112]. Most of the known functions of the VA RNAs were investigated using human Ad2 and Ad5 as model viruses. It was demonstrated that a virus mutant incapable of expressing VA RNAI suffers up to a 20-fold decrease in growth rate compare to the wild type virus [113]. Moreover when both VA RNAs are deleted, virus growth was reduced 60-fold [114]. However dele- tion of only VA RNAII had no effect on virus growth in tissue culture cells [113].

Figure 6. Schematic drawing showing the structure of adenovirus type 5 VA RNAI (left) and VA RNAII (right), adapted from [115].

A- VA RNA mediated-suppression of the innate immune response 1- VA RNA and PKR Initial clues about the function of the VA RNAs originated from a set of experiments conducted in Thomas Shenk’s laboratory, where they showed that late viral protein synthesis is dramatically reduced in cells infected with Ad5 mutants lacking VA RNAI [113]. The observed defects suggested that VA RNAI acts as a translational activator [116,117]. Absence of VA RNAI expression was shown to result in an increased phosphorylation of the α- subunit of the eukaryotic translation initiation factor 2 (eIF2α). Phosphory- lated elF2α has higher affinity for the guanine exchange factor, eIF-2B, which results in a failure to recharge eIF2 with GTP and therefore leads to a block in translation initiation [118]. Later on, it was demonstrated that activated protein kinase R (PKR) phosphorylates serine 51 in eIF2α, causing the late viral proteins translational arrest in the absence of VA RNAI [119,120]. PKR is a serine-threonine kinase that can be induced by interferon and be- comes active through binding to long dsRNA, usually produced during virus infections [121]. Binding of two PKR molecules to the same dsRNA results in a dimerization and activation of PKR through an autophosphorylation. In

28 the case of adenovirus, it has been suggested that the long dsRNA, which activates PKR is generated by viral DNA strands symmetrical transcription [122]. To ensure an efficient translation during an adenovirus infection, the apical stem of VA RNA strongly binds to PKR, while the central domain inhibit the PKR dimerization and consequently its activation [123].

Figure 7. Schematic drawing summarizing the inhibitory effects of VA RNAI on the interferon and the RNAi/miRNA pathways. Adapted from[115].

2- VA RNA and 2’-5’ Oligoadenylate Synthetase Another enzyme that can be induced by Interferon is 2’-5’ Oligoadenylate Synthetase (OAS). Similar to PKR, viral dsRNA causes an oligomerization of OAS to form a tetramer which will synthesize 2'-5'-oligoadenylates. 2'-5' oligoadenylates will in turn activate the latent form of ribonuclease L (RNaseL). Activated RNaseL will degrade both cellular and viral RNA in the infected cell, thus blocking virus growth and spreading [124]. In the case of an adenovirus infection, it has been shown that full length VA RNAI acts as a poor activator of the OAS enzyme, less than 10% compared to a perfect dsRNA of similar molecular weight [125]. On the other hand, it was demonstrated recently that a truncated VA RNAI, only containing the apical stem and central domain, has a much higher binding affinity to OAS than the full length VA RNAI. Interestingly this truncated version of VA RNAI functions as a pseudo substrate leading to inhibition of OAS activation. As will be described below, it is known that Dicer cleaves of the terminal stem of the VA RNA generating viral small RNAs, so called mivaRNA, leaving the

29 apical stem together with central domain acting as pseudo-substrate blocking OAS activation. Taken together these observations suggest that during an adenovirus infection, the Dicer-processed form of VA RNAI can outcompete the full length VA RNAI for OAS binding causing suppression of OAS activation [126].

B- VA RNA mediated suppression of the miRNA pathway As discussed above, extensive mutagenesis, structure probing and compara- tive sequence analysis demonstrated that the VA RNAs fold into conserved stem loop structures, which resembles cellular pre-miRNAs [112]. Similar to pre-miRNAs, VA RNAI export to the cytoplasm requires the Exportin 5 receptor [127]. This interaction is dependent on the double-stranded nature of the terminal stem and a 3 to 8 nucleotides overhang at the VA RNAI 3’ end. As a consequence VA RNAI will compete with cellular pre-miRNAs for Exportin 5 binding and transport to the cytoplasm. In fact it was demonstrated that overexpression of VA RNAI cause an accumulation of pre- miRNAs in the nucleus [127]. Moreover, the nuclear export of the Dicer mRNA is also mediated by Exportin 5, which is blocked in case of VA RNAI overexpression resulting in a reduction of the Dicer protein in Ad- infected cells [128]. Furthermore, VA RNA is involved in other molecular mechanisms that contribute to the suppression of the miRNA pathway. Once in the cytoplasm, Dicer processes both VA RNAI and VA RNAII into viral miRNAs, acting as competitive substrates lowering the Dicer capacity to cleave pre-miRNAs [55]. The terminal stems of VA RNAI and VA RNAII are cleaved into ≈ 22nts long small RNAs, so called mivaRNAI and mivaRNAII, respectively. During an adenovirus infection only a small portion of the total VA RNA pool (2-5%) is actually cleaved by Dicer [55,129]. However taking into account the large quantities of VA RNA produced during a lytic infection (>108 copies/cell), more than 106 mivaRNAs are generated at the late stage of an infection. Finally the processed mivaRNAs will be loaded onto Ago2 containing RISC complexes that are cleavage competent. The Ad2/5 mivaRNAs show highly asymmetric strand incorporation in RISC, with the 3’ strand of both VA RNAs (3’-mivaRNA) incorporated with more than a 200-fold higher efficiency compared to the 5’-strand (5’-mivaRNA). The 3’ strand of mivaRNAI generates unstable RISC complexes that have a dramatically reduced activity compared to the 5’ strand mivaRNAs [130]. Considering the massive production of the mivaRNAs late during an adenovirus infection, it was not unexpected that they would negatively influence the incorporation of cellular miRNAs to RISC as well. Taken all together, the adenovirus VA RNAs are the first, and so far the only, RNA-based RSS molecule encoded by a mammalian virus that is capable of inhibiting the miRNA/RNAi pathways at multiple levels.

30 Present investigation

Paper I: The adenovirus VA RNAI-derived miRNAs are not essential for lytic virus growth in tissue culture cells Besides blocking the miRNA pathway, it remained enigmatic whether the mivaRNAs possess a miRNA-like function, targeting cellular and/or viral mRNAs. Initial evidence about the miRNA-like function of mivaRNAs originated from a series of experiments demonstrating that the mivaRNAs are capable of binding to complementary sequence in the 3’UTR of reporter genes and inhibiting their expression [55]. Moreover it was shown that a fraction of the total mivaRNAs is associated with polyribosomes [129], which is commonly observed with functional cellular miRNA. Finally a large-scale microarray study showed that overexpression of the VA RNAs resulted in a significant down-regulation of 462 cellular mRNAs, most of these mRNAs were targets for the VA RNAI derived miRNAs [131]. These targets are involved in different cellular processes such as cell growth, transcription, RNA metabolism and DNA repair. However the key question re- mains whether these miRNA interactions are important for virus growth and that is what we wanted to address in this study.

Construction and characterization of recombinant adenoviruses with mutated mivaRNAI seed sequence VA RNA is a relatively small molecule that carries out several functions that are crucial for an adenovirus infection [115]. This fact makes it challenging to specifically investigate the function(s) of the processed mivaRNAs, even with the availability VA RNA deleted viruses. To circumvent this problem, we took advantage of the AdEasy system, where the adenovirus genome can easily be manipulated through homologous recombination. First we silenced the expression of the endogenous VA RNAI and II genes through mutation of the Box B of the RNA polymerase III internal promoters. By applying this methodology, we designed a new AdEasy virus vector backbone lacking functional VA RNAI and VA RNAII genes (pAdEasy-∆VA). After that the GFP gene was inserted into pAdEasy-∆VA creating pAdEasy-∆VA-GFP. Thereafter the GFP cassette in Ad-GFP-∆VA was replaced with the wild type VA RNAI gene, creating an adenovirus only capable of expressing VA RNAI. In addition, two more viruses were created were the 3’ and 5’ seed sequences, in mivaRNAI were changed through introduction of compensato-

31 ry mutations in the terminal stem of VA RNAI, creating the Ad-VAI 3’-mut and Ad-VAI 5’-mut, respectively. Based on the available literature, the introduced compensatory mutations do not change the VA RNAI secondary structure, and therefore should not interfere with other characterized VA RNA functions. Also our results showed that the mutated VA RNAIs were expressed as efficiently as the wild type counterparts. Intriguingly, through a manipulation of the mivaRNAI seed sequence, we managed to completely switch the strand selection properties of the mivaRNAIs in RISC during an adenovirus infection.

Mutation of the seed sequence in the 5’- or 3’-mivaRNAI does not perturb lytic adenovirus growth In order to investigate the effects of the introduced mutations on adenovirus growth, HEK293 cells were infected with the Ad-VAI wt or the Ad-VAI seed sequence mutated viruses or Ad-ΔVA-GFP for 22 hours followed by 35S pulse labelling for 2 hours. After that the production of late viral proteins was visualized both by autoradiography and western blotting, which measure the rate of synthesis and the steady state accumulation of viral proteins, respectively. The results from both assays indicated that viral late protein accumulation for the Ad-VAI seed sequence mutated viruses was comparable to the Ad-VAI wild type with no significant changes, while Ad-ΔVA-GFP failed to accumulate any detectable viral proteins. Moreover, the ability of a recombinant virus to proceed through a lytic infection and eventually causing cell death was measured via cell viability assay (MTS assay) at 4 and 6 days post infection. Ad-VAI 3’-mut and Ad-VAI 5’-mut showed essentially the same cytotoxicity as the recombinant virus expressing the wild type VA RNAI gene. On the other hand, the VA RNAI negative Ad-ΔVA-GFP exhib- ited severe growth inhibition with 20 folds difference compared to the Ad- VAI wt. The recombinant adenoviruses described in this study lack the endogenous adenovirus E1 region. Therefore, all experiments described so far were done in cells constitutively expressing the adenovirus E1 proteins: HEK293 or 911 cells, which support the growth of the recombinant viruses. However it was crucial to characterize these viruses in a cell line lacking expression of any adenovirus proteins such as Hela cells. To overcome this problem, we performed a trans-complementation assay. In this experiment we used a coinfection strategy with our seed-sequence mutant viruses and Ad5 mutant dl720, which contains an intact E1 region but lacks VA RNAI and VA RNAII expression [114]. Thus we investigated whether the recombinant viruses, expressing VA RNAI wt or seed mutants, could complement the dl720 growth, which in turn expresses E1 proteins supporting the growth of the recombinant viruses. The results demonstrated that the VA RNAI seed- mutant viruses fully complemented the VA RNAI function and supported growth of the dl720 virus. The functional complementation was demonstrat-

32 ed by the ability of the defective dl720 mutant virus to proceed to the late phase of a lytic infection causing both production of late viral proteins and shut-off of host cell protein translation.

Paper II: Small RNA sequence analysis of adenovirus VA RNA-derived miRNAs reveals an unexpected serotype-specific difference in structure and abundance In this paper we investigated the function of the adenovirus-encoded mivaRNAs by an alternative approach. By selecting a representative serotype from five different adenovirus species, B2 (Ad11), C (Ad5), D (Ad37), and E (Ad4), we investigated whether the VA RNAs expressed from different adenovirus species undergo Dicer-mediated processing into mivaRNAs, which will be subsequently loaded into RISC. Also the study gave us in- sights to whether the mivaRNAs are similar in sequence and abundance to the mivaRNAs expressed from the well-characterized human Ad5. Such information can shed some light on the evolutionary conserved function(s) of the adenovirus mivaRNAs.

The impact of different adenovirus serotypes infection on the miRNA- pathway components Different adenovirus serotypes show significant genome diversity, which affects different aspect of the virus life cycle, like cell entry, evasion of the immune system, and the mechanism involved in the hijacking of host cell pathways. To monitor the progress of viral growth Hela cells were infected with Ad4, Ad5, Ad11 and Ad37 followed by a 35S-methionine labelling for 2 hours at 24 and 48 hpi. The result showed that Ad4, Ad5 and Ad37 showed the fastest growth rate as visualized by the production of late viral proteins and the shut-off of host cell protein synthesis at 24 hpi. Later on, viral protein production decreased for Ad5 and Ad37, while it completely stopped in Ad4 infected cells. On the other hand, Ad11 showed a slower growth pheno- type and the production of late viral protein became distinct only at 48 hpi. To examine the interaction between different adenovirus serotypes and the miRNA pathway, we investigated the impact of the adenovirus infections on the steady state levels of the main RNAi components; Exportin 5, respon- sible for export pre-miRNA from the nucleus into cytoplasm; Dicer, which process short hairpin into small RNA duplexes; TRBP, the Dicer partner protein needed for efficient miRNA processing; and finally Ago2 a core component of RISC, that binds to small RNA and mediates gene silencing. The result showed that in Hela cells infected with different serotypes the steady state levels of Exportin 5, Dicer, TRBP and Ago2 were not signifi-

33 cantly changed at 24 hpi. However, at 48 hpi the levels of Dicer and TRBP were reduced selectively in Ad4 infected cells.

Accumulation of VA RNAI and VA RNAII during an adenovirus infection VA RNAI and II sequences among different adenovirus species are not well conserved. However all fold into a pre-miRNA like structure. Previous results have shown that the Ad5 VA RNAI and II are processed into small RNAs 22nts long. Here we investigated whether such a mechanism is conserved between members of different species. To address this question we infected Hela cells with the Ad4, Ad5, and Ad11 or Ad37 viruses. At 24 hpi total RNA was extracted followed by Northern blotting with radiolabeled probes detecting the 3’ or the 5’ stems of the VA RNAs. As predicted, the Northern blot results showed that VA RNAI from the different species is processed into small RNA of about 22ntss. Moreover, the band corresponding to small RNA generated from the 3’ stem of VA RNAI (3’-mivaRNAI) were more distinct than the one generated from 5’ end of VA RNAI (5’- mivaRNAI) for all tested serotypes. The Ad11 genome contains only one VA RNA gene, so the investigation of VA RNAII processing was only performed for serotypes Ad4, Ad5 and Ad37. In agreement with a previous report on Ad5 [129], the 3’-mivaRNAII was most distinct. A similar pattern was observed for Ad4 VA RNAII. Intri- guingly both the 3’- and the 5’-mivaRNAII from Ad37 VARNAII were detected as sharp bands around 22nts by Northern blotting.

Dicer-dependent processing of the adenovirus small RNAs Because of the secondary structure configuration of the VA RNAs, it ap- peared likely that the mivaRNAs are Dicer processing products. This has been proven to be true in case of the Ad5 VA RNAI and II small RNAs, based on in vitro Dicer-cleavage assay [55]. Since our results demonstrated that all VA RNAs are processed into small RNAs, we were curious to know whether the Dicer-processing step is conserved among all adenovirus species. To test this we used Dicer-specific siRNAs to knockdown Dicer protein expression in Hela cells. After 48 hours of siRNA treatment cells were infected with the four serotypes used in this study. At 24 hpi the accumulation of the 3’ stem of both VA RNAI and II was detected using a Northern blot assay. We analyzed 3’-mivaRNA expression since this small RNA was preferentially accumulating in all of the tested serotypes. The results showed that the accumulations of the 3’-mivaRNAI and 3’-mivaRNAII from all serotypes were significantly reduced after reduced Dicer protein expression. Also in case of Dicer knockdown, there was an increase in the full length VA RNAI and VA RNAII accumulation accompanied with appearance of other RNA species, which are likely to be VA RNA processing intermedi- ates.

34 Incorporation of the processed mivaRNAI and mivaRNAII into RISC The last step in the maturation of miRNA is the association with the Ago proteins to form RISC. The newly formed miRISC will then target complementary mRNAs leading to their down regulation. To investigate whether these newly identified mivaRNAs can associate with Ago2, the main component of RISC, we infected 293-Flag-Ago2 cells [129], which constitutively express a Flag-epitope tagged Ago2 protein with the four adenovirus serotypes. After 24 hpi, the Ago2 protein was immunopurified using an anti-Flag resin and associated small RNAs extracted. For VA RNAI a preference for Ago2 incorporation of the 3’-mivaRNAI was observed for all tested serotypes. Interestingly the 5’-mivaRNAI from Ad11 was incorporated into RISC significantly more than other serotypes. For VA RNAII a similar strand preference for the 3’-mivaRNAIIs was observed. Interestingly in case of Ad37 both the 3’- and 5’-mivaRNAII were significantly more incorporated in RISC compared to other serotypes.

Cloning of cytoplasmic and RISC-associated small RNAs in Ad-infected cells In order to investigate in depth the small RNA production in the different serotypes, we sequenced both the total cytoplasmic and RISC associated small RNAs from Ad4, Ad5, Ad11, Ad37 and mock infected 293-Flag-Ago2 cells at 24 hpi. In the present study we focused on characterization of the newly identified mivaRNAs and any other small RNAs generated from the adenovirus genomes. Thus, reads obtained from deep sequencing were mapped against the corresponding adenovirus genomes (Ad4, Ad5, Ad11 or Ad37). For all four serotypes small RNAs were mainly generated from the VA RNA region, specifically from the 5’ and 3’ arms of the terminal stem. - Moreover there was a clear enrichment in the reads generated from the VA RNA region in the RISC fraction compared to the cytoplasmic fraction. In agreement with the Northern blot analysis, Ad11 and Ad37 showed the most dramatic increase in RISC association of mivaRNAs in 293-Flag-Ago2 cells. The read size distribution of the mapped small RNAs was narrowed around 19-24 nts in the RISC fraction compared to a more broad distribution in the cytoplasmic fraction, which indicates the successful enrichment of the small RNA in our RISC immunoprecipitation experiment. Thus further analysis was focused on this enriched small RNA species ranging from size 19 to 24 nucleotides. Analysis of our sequencing data suggests that the 3’-mivaRNAI and 3’- mivaRNAII generated from the different serotypes, except for Ad37 VAR- NAII, are preferentially accumulating compared to the 5’-mivaRNAI and 5’- mivaRNAII. Also the 3’-mivaRNAII from all serotypes seemed to be significantly enriched in the RISC fraction. For instance Ad4 mivaRNAII-147, Ad5 mivaRNAII-136 and Ad37 mivaRNAII-128 were enriched from 5%,

35 1% and 3.4% in the cytoplasmic fraction to 13.9%, 15.5% and 18.5% in the RISC fraction, respectively. This enrichment was not as apparent in the case of the 3’-mivaRNAI produced by the tested adenovirus serotypes.

Paper III: Characterization of a non-canonical 5’ capped adenoviral small RNA that is functionally active There are several studies that have utilized high throughput sequencing methods to identify viral small RNA that are produced during an Ad5 lytic infection [129,132,133] and persistent/latent infection [134]. Collectively these studies have suggested that the VA RNA gene(s) most likely is the only source for functional small RNA produced during an adenovirus infection. Our recent serotype specific cloning experiments (Paper II) corroborat- ed these studies by showing that the mivaRNAs are the major viral small RNAs produced. However, a deeper analysis of this data suggested that additional, less abundant, small RNAs were enriched in Ago2 containing RISC. Here we characterize a second abundant viral small RNA class produced during an adenovirus infection, the so-called transcriptional start site (TSS) small RNAs.

Production of RISC associated m7G-capped small RNAs from the Ad37 MLP transcriptional start site A closer analysis of the data from Ad37 infected cells demonstrated that a new class of small RNAs accumulated to significant levels. These RNAs mapped downstream of the adenoviral MLP, sharing the transcription start site (TSS) as the rest of the major late transcripts. The majority of these small RNAs starts at the +1 position and are 31 nucleotides long, designated as MLP-TSS-sRNA (+1). A minor species of the TSS-sRNA start three nucleotides upstream of the major late cap site and are 34 nucleotides long. These are designated as MLP-TSS-sRNA (-3). Since the 5’end of the MLP-TSS-sRNA coincides with the characterized cap site of the MLP, we were curious to investigate whether the MLP-TSS- sRNAs were m7G capped. Experiment done on cytoplasmic RNA from infected cells treated with enzymes capable of removing cap structure from the RNA 5’end (Tobacco acid pyrophosphatase (TAP) and RNA 5' Pyrophos- phohydrolase (RppH)) demonstrated that enzymatic treatment significantly enhanced the crosslinking efficiency of the MLP-TSS-sRNA to the membrane, suggesting that they were m7G-capped. Importantly, we showed that the MLP-TSS-sRNA were associated with Ago2 containing RISC. In addition, an estimation of the Ago2-bound and unbound small RNA fractions suggested that the majority of MLP-TSS-sRNAs in RISC are capped.

36 The MLP-TSS-sRNAs are produced by a non-canonical pathway Moreover, we showed that unlike the mivaRNAs, the MLP-TSS-sRNAs are produced by another pathway, not requiring the Dicer enzyme or the Ex- portin5 export factor. Since the MLP-TSS-sRNAs binds to Ago2 complexes, we were curious to investigate whether Ago2 could be involved in MLP- TSS-sRNA maturation, similar to the cellular miR-451miRNA. Interestingly we found that MLP-TSS-sRNA binds both the catalytically dead and wild type Ago2 protein with similar efficiencies during an Ad37 infection, indicating the Ago2 slicer activity is not required for MLP-TSS-sRNA processing.

MLP-TSS-RNAs downregulate target mRNAs with a complementary sequence Since the MLP-TSS-sRNAs differ in many features from the rest of the Ago- bound small RNA species, it became of interest to investigate whether the MLP-TSS-sRNA can still induce post-transcriptional gene silencing of a target mRNA. For this experiment reporter plasmids with the MLP-TSS- sRNA target sequences where generated. The result from both Ad37 infections and transient MLP-TSS-sRNA transfections suggested that the MLP- TSS-sRNAs are capable of sequence specifically suppress reporter mRNAs having complementary binding sites in their 3’UTR.

MLP-TSS-RNAs suppresses an Ad37 infection However the most important question to answer is how the MLP-TSS- sRNAs affect a normal Ad37 infection? Interestingly, the MLP-TSS-sRNA originates from the opposite strand, with prefect sequence complementary binding sites to the E2B transcription unit, which encodes for the essential viral replication factors, AdPol and pTP. To test the possibility that MLP- TSS-sRNA can target the AdPol and pTP mRNAs, we transfected the MLP- TSS-sRNA or a scrambled single stranded RNA (ssRNA) and followed their effect on the progression of an Ad37 infection. The results were intriguing showing that both viral DNA copy number and capsid protein synthesis were significantly reduced after MLP-TSS-sRNA transfection compared to the scrambled ssRNA, indicating that MLP-TSS-sRNAs, indeed, limits viral DNA replication.

37 Conclusions and future perspectives

In paper I, the results pointed to the non-essential role of the human Ad5- mivaRNAIs seed sequences for the lytic virus life cycle. This conclusion favours the model where VA RNA mainly act as a suppressor of the miRNA pathway and antiviral interferon machineries, represented by PKR and OAS. However we should also take into account that this study was performed in standard cell culture systems where lytic virus growth occurs very rapidly and requires as little as 30 hours for a complete replication cycle. Thus, this time span is, most likely, to short to detect a potential impact of mivaRNA regulation of gene expression on the adenovirus life cycle. Such effects will be more relevant in a more slow natural acute infection in humans. However that would be extremely difficult to investigate, since human adenovirus replicates very poorly in current animal models [135]. An alternative strategy would be to expand the investigation of mivaRNA function to other cell lines, like T and B cells, where adenovirus potentially can establish persistent/long term infections that may mimic the in vivo situation [136,137]. In paper II, we showed that the Dicer-dependent production of the mivaRNAs is a conserved mechanism shared by several adenovirus serotypes, suggesting that they play, a yet to be identified, important function in the adenovirus life cycle. Interestingly Ad11 and Ad37 accumulate higher levels of the viral small RNAs in comparison with the well-characterized Ad5, suggesting that the role-played by the mivaRNAs might be more important in these serotypes. To get a deeper insight into mivaRNA function, it would be interesting to create mivaRNA-seed mutant viruses for both Ad11 and Ad37. However, that could be technically challenging since established viral vector systems have not been described for these serotypes. Another approach would be to identify direct mRNA targets of the mivaRNAs through applying High-throughput sequencing of RNA isolated by crosslinking immunoprecipitation (HITS-CLIP) technique from RISC-purified complexes in Ad11 and Ad37 infected cells. In paper III, the discovery of the MLP-TSS-sRNA added a whole new chapter to the adenoviral small RNA story. It demonstrates an example in which a mammalian virus utilizes two different routes for production of viral small RNAs. Since VARNA is a global suppressor of the canonical miRNA pathway it will not only compete with cellular miRNA production but also any other potential viral small RNA, dependent on Dicer and Exportin5 for its generation. Therefore to avoid this potential conflict, adenovirus utilizes a

38 novel route, yet to be characterized, for the generation of the MLP-TSS- sRNA. The MLP-TSS-sRNA seems to be involved in regulation of the switch from the early to the late phase of infection (the demarcation line is the initiation of viral DNA replication) through targeting the expression of the adenoviral E2B replication proteins (Adpol and pTP). Thus, MLP-TSS- sRNA overexpression suppresses the efficiency of viral DNA replication. Among the questions that need to be addressed is whether the MLP-TSS- sRNA can also regulate cellular genes. To answer this question, we will utilize the approach mention above, HITS-CLIP, to identify MLP-TSS-sRNA targets in Ad37 infected cells or synthetic MLP-TSS-sRNA transfected cells. Also it would be intriguing to investigate if there are cellular small RNAs with similar features as the discovery of the MLP-TSS-sRNA. To answer this question we will use the double-immunoprecipitation protocol described in this paper (Ago2 + m7G-cap pull down) to enrich for Ago2-associated capped small RNAs, followed by high throughput sequencing. Finally the discovery of the MLP-TSS-sRNA is not just important for adenovirus biology but may also have a role in gene therapy-mediated by small RNA. Since the MLP-TSS-sRNA functions as 5’capped single stranded RNA in transient transfection assays it suggests that it should be possible to engineer similar molecules with different 5’ modification that can enhance the stability or delivery of therapeutic sRNAs.

39 Acknowledgment

I would like to start this section by saying thanks to my supervisor and men- tor professor Göran Akusjärvi. Thank you Göran! For giving me the op- portunity to do my PhD in your group, for the freedom to explore both worlds of non-coding RNA and virology, for your guidance, support and being always available for discussions and questions. I really admire your passion to educate your students and your post-docs in any available mo- ment. Also I am really glad to be your 25th PhD student (that must be worth celebrating, right! J).

Tanel, my co-supervisor and co-author, thanks for the technical advises and your continuous positive altitude toward my results, we had a fruitful collab- oration, which I really hope it will continue in the future. Special thanks to Catherine for sharing your enormous knowledge in molecular biology and virology during Friday meetings. Göran Magnusson for constructive criti- cism during the years and being the chairman of my PhD defense committee. My dear friend Roger Chang many thanks for the nice long discussions about science and life in general, good luck in your PhD!

Shady, my Egyptian friend in IMBIM, we both came to Sweden as master students and started our PhD in about the same time, good time flies quickly.

Ravi, thanks for your 24/7 cheerful mode, and our very long talks about Indian culture!

Heidi, my favorite office mate, thank you for sharing your enthusiasm and positive energy. Ellenor and Cecilia thanks for your support and encour- agement. Anette I am deeply grateful for all help during all these years, and for always finding time to answer my questions. Xin, I enjoyed your talk about your business plans, good luck in future. To all the past and present members of Adeno group and IMBIM for such great time.

My Egyptian friends here in Uppsala Farag, Amr, Mohammed, Ahmed, Shady and Assem, thank you for the wonderful time and wish you all the best.

40 My older brothers, Tamer and Mohammed, Thank you for always setting up an example I can look up to since we were kids.

My dear wife, Fatma, what can I say!! Thanks for the constant and over- whelming love, patience and endless support. If this thesis and the work described in it are of any good, it is because of you! I must be very lucky and blessed because I met you.

My son, Yassin, your smile and laugh were more than enough to forget all the frustration from the lab and failed experiments, and helped me to move forward.

My Dad and mom, Thank you for everything, for your sacrifice, support and unconditional love…

اﻟﻤﻠﺨﺺ اﻟﻌﺮﺑﻲ

ﻣﻨﺬ أن اﻛﺘﺸﻔﺖ اﻟﻔﯿﺮوﺳﺎت اﻟﻐﺪاﻧﯿﺔ (Adenoviruses) ﻓﻰ ﻋﺎم 1953 وﻗﺪ ﻇﻬﺮ دورﻫﺎ اﻟﻬﺎم ﻓﻰ ﻓﻬﻢ اﻟﺘﻔﺎﻋﻼت داﺧﻞ اﻟﺨﻠﯿﺔ اﻟﻤﻀﯿﻔﺔ ﻟﻠﻔﯿﺮوس واﻟﺘﻰ ﺗﺸﻤﻞ اﻟﺘﻔﺎﺻﯿﻞ اﻟﻤﯿﻜﺎﻧﺰﻣﯿﺔ ﻟﻠﻌﻤﻠﯿﺎت اﻟﺨﻠﻮﯾﺔ ﻣﺜﻞ اﻟﺘﺤﻜﻢ ﻓﻰ دورة اﻟﺨﻠﯿﺔ واﻟﺮﺑﻂ اﻟﻤﺘﺒﺎدل ﻟﻠﺤﻤﺾ اﻟﻨﻮوى اﻟﺮﯾﺒﻮزى .وﻣﻦ أواﺋﻞ اﻟﺠﯿﻨﺎت اﻟﻔﯿﺮوﺳﯿﺔ اﻟﺘﻰ ﺗﻢ وﺻﻔﻬﺎ ودراﺳﺘﻬﺎ ﻟﻬﺬه اﻟﻔﯿﺮوﺳﺎت ﻫﻢ ﺟﯿﻨﺎت اﻟﺤﻤﺾ اﻟﺮﯾﺒﻮزى اﻟﻤﺮﺗﺒﻂ ﺑﺎﻟﻔﯿﺮوس (Viral associated RNA , VA RNAI/II), واﻟﺘﻰ ﺗﻨﺘﺞ ﺑﻜﻤﯿﺎت ﻫﺎﺋﻠﺔ أﺛﻨﺎء اﻟﻌﺪوى اﻟﺘﺤﻠﻠﻲ. ﻫﺬه اﻟﺠﯿﻨﺎت ﺗﺆدى اﻟﻜﺜﯿﺮ ﻣﻦ اﻟﻮﻇﺎﺋﻒ اﻟﻤﻬﻤﺔ ﻹﺗﻤﺎم دورة ﺣﯿﺎة ﻧﺎﺟﺤﺔ ﻟﻠﻔﯿﺮوس. و ﺷﻮﻫﺪ ﺣﺪﯾﺜﺎ أن VA RNA ﯾﻘﻄﻊ إﻟﻰ أﺣﻤﺎض رﯾﺒﻮزﯾﺔ ﺻﻐﯿﺮة ﺗﺴﻤﻰ ال mivaRNAs, واﻟﺘﻰ ﺗﺘﻌﺎرض ﻣﻊ وﻇﺎﺋﻒ وآﻟﯿﺔ ال RNAi/miRNA pathway. ﻓﻰ اﻟﺒﺤﺚ اﻷول واﻟﺜﺎﻧﻰ, ﺗﻢ اﻟﺘﺮﻛﯿﺰ ﻋﻠﻰ اﻟﻮﺻﻒ اﻟﺘﺮﻛﯿﺒﻰ واﻟﻮﻇﯿﻔﻰ ﻻل mivaRNAs ﺑﺈﺳﺘﺨﺪام ﻃﺮﯾﻘﺘﯿﻦ .ﻓﻰ اﻟﻄﺮﯾﻘﺔ اﻷوﻟﻰ ﺗﻢ اﺑﺘﻜﺎر ﻧﻈﺎم ﻧﻤﻮذﺟﻰ ﯾﻤﻜﻨﻨﺎ ﻣﻦ دراﺳﺔ وﻇﺎﺋﻒ ال mivaRNAI ﺑﺪون اﻟﻤﺴﺎس ﺑﺎﻟﻮﻇﺎﺋﻒ اﻷﺧﺮى ل VA RNA . وﻗﺪ ﺗﺤﻘﻖ ذﻟﻚ ﻣﻦ ﺧﻼل ﺑﻨﺎء ﻓﯿﺮوﺳﺎت ﻏﺪاﻧﯿﺔ ﻣﻌﺪﻟﺔ ﺑﻄﻔﺮات ﻓﻰ اﻟﺘﺴﻠﺴﻞ اﻟﺒﺬرى(Seed Sequence) ﻟﻞ mivaRNAI. وﻗﺪ أﻇﻬﺮت ﻧﺘﺎﺋﺞ اﻟﺘﺠﺎرب ﻓﻰ اﻟﺨﻼﯾﺎ اﻟﺒﺸﺮﯾﺔ أن ﻫﺬة اﻟﻔﯿﺮوﺳﺎت اﻟﻐﺪاﻧﯿﺔ اﻟﻤﻌﺪﻟﺔ وراﺛﯿﺎً ﻧﻤﺖ ﻛﺎﻟﻨﻮع اﻟﺒﺮى ﻣﻤﺎ أﺷﺎر إﻟﻰ أن mivaRNAs ﻏﯿﺮ ﻣﻬﻤﺔ ﻓﻰ اﻟﻤﺮﺣﻠﺔ اﻟﺘﺤﻠﻠﯿﺔ ﻣﻦ اﻟﻌﺪوى ﺑﺎﻟﻔﯿﺮوس. وﻓﻰ اﻟﺒﺤﺚ اﻟﺜﺎﻧﯿﺔ وﺟﺪﻧﺎ أن VA RNA اﻟﺬى ﯾﺘﻢ اﻧﺘﺎﺟﺔ ﻣﻦ اﻟﻔﯿﺮوﺳﺎت اﻟﻐﺪاﻧﯿﺔ اﻟﻤﺨﺘﻠﻔﺔ (أد4, أد5, أد11, أد37) ﯾﺨﻀﻊ ﻟﻨﻔﺲ ﻧﻮع اﻟﻤﻌﺎﻟﺠﺔ إﻟﻰ mivaRNAs واﻟﺘﻰ ﻵﺣﻘﺎ ﺗﺤﻤﻞ ﺑﻜﺎﻓﺎﺋﺎت ﻣﺨﺘﻠﻔﺔ ﻋﻠﻰ RISC (اﻟﻤﺮﻛﺐ اﻟﻤﺜﺒﻂ اﻟﻤﻮﺟﺔ ﺑﺎﻟﺤﻤﺾ اﻟﺮﯾﺒﻮزى RNA-induced silencing complex). وﻓﻰ اﻟﺒﺤﺚ اﻟﺜﺎﻟﺚ, أﺛﺒﺘﻨﺎ أن اﻟﻤﻨﻄﻘﺔ اﻟﻘﺮﯾﺒﺔ ﻣﻦ MLP ﺗﻨﺘﺞ ﻃﺒﻘﺔ ﻏﯿﺮ ﻣﺘﻌﺎرف ﻋﻠﯿﻬﺎ ﻣﻦ اﻷﺣﻤﺎض اﻟﺮﯾﺒﻮزﯾﺔ اﻟﺼﻐﯿﺮة واﻟﺘﻰ ﺳﻤﯿﺖ ب MLP-TSS-sRNAs. واﻟﻤﺜﯿﺮ ﻟﻠﺪﻫﺸﺔ ﻫﻮ أن ﻫﺬا اﻟﺤﻤﺾ اﻟﺮﯾﺒﻮزي اﻟﺼﻐﯿﺮMLP-TSS-sRNAs ﯾﺤﺎﻓﻆ ﻋﻠﻰ m7G-cap أﺛﻨﺎء إرﺗﺒﺎﻃﻪ ﺑﺎﻟﻤﺮﻛﺐ اﻟﻤﺜﺒﻂ RISC. وﻫﺬه اﻟﻤﺮﻛﺒﺎت اﻟﻤﺤﻤﻠﺔ ب MLP-TSS- sRNAs ﻗﺎدرة ﻋﻠﻰ اﺳﺘﻬﺪاف اﻷﺣﻤﺎض اﻟﺮﯾﺒﻮزﯾﺔ ذات أﻣﺎﻛﻦ رﺑﻂ ﻣﻜﻤﻠﺔ ( complementary binding site). واﻷﻛﺜﺮ أﻫﻤﯿﺔ أن MLP-TSS-sRNA ﯾﻘﻠﻞ ﻣﻦ ﻛﻔﺎءة ﺗﻀﺎﻋﻒ اﻟﺤﻤﺾ اﻟﻨﻮوى اﻟﻔﯿﺮوﺳﻰ وذﻟﻚ ﻣﻦ ﺧﻼل إﺳﺘﻬﺪاف اﻷﺣﻤﺎض اﻟﺮﯾﺒﻮزﯾﺔ اﻟﻔﯿﺮوﺳﯿﺔ E2B mRNAs. وﻓﻰ اﻟﺨﺘﺎم, MLP-TSS-sRNA ﯾﻤﺜﻞ أول ﺣﻤﺾ رﯾﺒﻮزى ﻓﯿﺮوﺳﻰ ﺻﻐﯿﺮ ﺷﻮﻫﺪ أن ﻟﻪ وﻇﯿﻔﺔ ﻛﻤﻨﻈﻢ ﻟﻌﺪوى اﻟﻔﯿﺮوﺳﺎت اﻟﻐﺪاﻧﯿﺔ.

42 References

1. Jacob F, Monod J (1961) Genetic regulatory mechanisms in the synthesis of proteins. J Mol Biol 3: 318-356. 2. Weinberg RA, Penman S (1968) Small molecular weight monodisperse nuclear RNA. J Mol Biol 38: 289-304. 3. Dreyfuss G, Philipson L, Mattaj IW (1988) Ribonucleoprotein particles in cellular processes. J Cell Biol 106: 1419-1425. 4. Maxwell ES, Fournier MJ (1995) The small nucleolar RNAs. Annu Rev Biochem 64: 897-934. 5. Meier UT (2005) The many facets of H/ACA ribonucleoproteins. Chromosoma 114: 1-14. 6. Kruger K, Grabowski PJ, Zaug AJ, Sands J, Gottschling DE, et al. (1982) Self- splicing RNA: autoexcision and autocyclization of the ribosomal RNA intervening sequence of Tetrahymena. Cell 31: 147-157. 7. Guerrier-Takada C, Gardiner K, Marsh T, Pace N, Altman S (1983) The RNA moiety of ribonuclease P is the catalytic subunit of the enzyme. Cell 35: 849- 857. 8. Laylin JK (1993) Nobel laureates in chemistry, 1901-1992: Chemical Heritage Foundation. 9. Okazaki Y, Furuno M, Kasukawa T, Adachi J, Bono H, et al. (2002) Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs. Nature 420: 563-573. 10. Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, et al. (2005) The transcriptional landscape of the mammalian genome. Science 309: 1559-1563. 11. Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, et al. (2005) Antisense transcription in the mammalian transcriptome. Science 309: 1564- 1566. 12. Djebali S, Davis CA, Merkel A, Dobin A, Lassmann T, et al. (2012) Landscape of transcription in human cells. Nature 489: 101-108. 13. Liu G, Mattick JS, Taft RJ (2013) A meta-analysis of the genomic and transcriptomic composition of complex life. Cell Cycle 12: 2061-2072. 14. Wightman B, Ha I, Ruvkun G (1993) Posttranscriptional regulation of the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in C. elegans. Cell 75: 855-862. 15. Braun L, Cannella D, Ortet P, Barakat M, Sautel CF, et al. (2010) A complex small RNA repertoire is generated by a plant/fungal-like machinery and effected by a metazoan-like Argonaute in the single-cell human parasite Toxoplasma gondii. PLoS Pathog 6: e1000920. 16. Missal K, Rose D, Stadler PF (2005) Non-coding RNAs in Ciona intestinalis. Bioinformatics 21 Suppl 2: ii77-78. 17. Shi W, Hendrix D, Levine M, Haley B (2009) A distinct class of small RNAs arises from pre-miRNA-proximal regions in a simple chordate. Nat Struct Mol Biol 16: 183-189.

43 18. Carnero E, Sutherland JD, Fortes P (2011) Adenovirus and miRNAs. Biochim Biophys Acta 1809: 660-667. 19. Kincaid RP, Sullivan CS (2012) Virus-Encoded microRNAs: An Overview and a Look to the Future. PLoS Pathog 8: e1003018. 20. Kozomara A, Griffiths-Jones S (2011) miRBase: integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res 39: D152-157. 21. Cai X, Hagedorn CH, Cullen BR (2004) Human microRNAs are processed from capped, polyadenylated transcripts that can also function as mRNAs. RNA 10: 1957-1966. 22. Lee Y, Kim M, Han J, Yeom KH, Lee S, et al. (2004) MicroRNA genes are transcribed by RNA polymerase II. EMBO J 23: 4051-4060. 23. Borchert GM, Lanier W, Davidson BL (2006) RNA polymerase III transcribes human microRNAs. Nat Struct Mol Biol 13: 1097-1101. 24. Hutvagner G, McLachlan J, Pasquinelli AE, Balint E, Tuschl T, et al. (2001) A cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA. Science 293: 834-838. 25. Lee Y, Ahn C, Han J, Choi H, Kim J, et al. (2003) The nuclear RNase III Drosha initiates microRNA processing. Nature 425: 415-419. 26. Yi R, Qin Y, Macara IG, Cullen BR (2003) Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev 17: 3011-3016. 27. Peters L, Meister G (2007) Argonaute proteins: mediators of RNA silencing. Mol Cell 26: 611-623. 28. Ruby JG, Jan CH, Bartel DP (2007) Intronic microRNA precursors that bypass Drosha processing. Nature 448: 83-86. 29. Okamura K, Hagen JW, Duan H, Tyler DM, Lai EC (2007) The mirtron pathway generates microRNA-class regulatory RNAs in Drosophila. Cell 130: 89-100. 30. Berezikov E, Chung WJ, Willis J, Cuppen E, Lai EC (2007) Mammalian mirtron genes. Mol Cell 28: 328-336. 31. Zamudio JR, Kelly TJ, Sharp PA (2014) Argonaute-bound small RNAs from promoter-proximal RNA polymerase II. Cell 156: 920-934. 32. Xie M, Li M, Vilborg A, Lee N, Shu MD, et al. (2013) Mammalian 5'-capped microRNA precursors that generate a single microRNA. Cell 155: 1568-1580. 33. Cifuentes D, Xue H, Taylor DW, Patnode H, Mishima Y, et al. (2010) A novel miRNA processing pathway independent of Dicer requires Argonaute2 catalytic activity. Science 328: 1694-1698. 34. Cheloufi S, Dos Santos CO, Chong MM, Hannon GJ (2010) A dicer- independent miRNA biogenesis pathway that requires Ago catalysis. Nature 465: 584-589. 35. Yoda M, Cifuentes D, Izumi N, Sakaguchi Y, Suzuki T, et al. (2013) Poly(A)- specific ribonuclease mediates 3'-end trimming of Argonaute2-cleaved precursor microRNAs. Cell Rep 5: 715-726. 36. Xie MY, Steitz JA (2014) Versatile microRNA biogenesis in animals and their viruses. Rna Biology 11: 673-681. 37. Lee RC, Feinbaum RL, Ambros V (1993) The C. elegans heterochronic gene lin- 4 encodes small RNAs with antisense complementarity to lin-14. Cell 75: 843- 854. 38. Filipowicz W, Bhattacharyya SN, Sonenberg N (2008) Mechanisms of post- transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9: 102-114. 39. Vella MC, Choi EY, Lin SY, Reinert K, Slack FJ (2004) The C. elegans microRNA let-7 binds to imperfect let-7 complementary sites from the lin-41 3'UTR. Genes Dev 18: 132-137.

44 40. Brennecke J, Stark A, Russell RB, Cohen SM (2005) Principles of microRNA- target recognition. PLoS Biol 3: e85. 41. Friedman RC, Farh KKH, Burge CB, Bartel DP (2009) Most mammalian mRNAs are conserved targets of microRNAs. Genome Research 19: 92-105. 42. Baltimore D (1971) Expression of animal virus genomes. Bacteriol Rev 35: 235- 241. 43. Both GW, Furuichi Y, Muthukrishnan S, Shatkin AJ (1975) Ribosome binding to reovirus mRNA in protein synthesis requires 5' terminal 7-methylguanosine. Cell 6: 185-195. 44. Both GW, Banerjee AK, Shatkin AJ (1975) Methylation-dependent translation of viral messenger RNAs in vitro. Proc Natl Acad Sci U S A 72: 1189-1193. 45. Muthukrishnan S, Both GW, Furuichi Y, Shatkin AJ (1975) 5'-Terminal 7- methylguanosine in eukaryotic mRNA is required for translation. Nature 255: 33-37. 46. Furuichi Y, Morgan M, Muthukrishnan S, Shatkin AJ (1975) Reovirus messenger RNA contains a methylated, blocked 5'-terminal structure: m- 7G(5')ppp(5')G-MpCp. Proc Natl Acad Sci U S A 72: 362-366. 47. Sharp PA (1994) Split genes and RNA splicing. Cell 77: 805-815. 48. Berget SM, Moore C, Sharp PA (1977) Spliced Segments at 5' Terminus of Adenovirus 2 Late Messenger-Rna. Proceedings of the National Academy of Sciences of the United States of America 74: 3171-3175. 49. Chow LT, Gelinas RE, Broker TR, Roberts RJ (1977) An amazing sequence arrangement at the 5' ends of adenovirus 2 messenger RNA. Cell 12: 1-8. 50. Sedman SA, Mertz JE (1988) Mechanisms of synthesis of virion proteins from the functionally bigenic late mRNAs of simian virus 40. J Virol 62: 954-961. 51. Lagos-Quintana M, Rauhut R, Lendeckel W, Tuschl T (2001) Identification of novel genes coding for small expressed RNAs. Science 294: 853-858. 52. Lee RC, Ambros V (2001) An extensive class of small RNAs in Caenorhabditis elegans. Science 294: 862-864. 53. Lau NC, Lim LP, Weinstein EG, Bartel DP (2001) An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science 294: 858-862. 54. Pfeffer S, Zavolan M, Grasser FA, Chien M, Russo JJ, et al. (2004) Identification of virus-encoded microRNAs. Science 304: 734-736. 55. Andersson MG, Haasnoot PC, Xu N, Berenjian S, Berkhout B, et al. (2005) Suppression of RNA interference by adenovirus virus-associated RNA. J Virol 79: 9556-9565. 56. Cai X, Lu S, Zhang Z, Gonzalez CM, Damania B, et al. (2005) Kaposi's sarcoma-associated herpesvirus expresses an array of viral microRNAs in latently infected cells. Proc Natl Acad Sci U S A 102: 5570-5575. 57. Grey F, Antoniewicz A, Allen E, Saugstad J, McShea A, et al. (2005) Identification and characterization of human cytomegalovirus-encoded microRNAs. J Virol 79: 12095-12099. 58. Samols MA, Hu J, Skalsky RL, Renne R (2005) Cloning and identification of a microRNA cluster within the latency-associated region of Kaposi's sarcoma- associated herpesvirus. J Virol 79: 9301-9305. 59. Bogerd HP, Karnowski HW, Cai X, Shin J, Pohlers M, et al. (2010) A mammalian herpesvirus uses noncanonical expression and processing mechanisms to generate viral MicroRNAs. Mol Cell 37: 135-142. 60. Cazalla D, Xie M, Steitz JA (2011) A primate herpesvirus uses the integrator complex to generate viral microRNAs. Mol Cell 43: 982-992.

45 61. Kincaid RP, Burke JM, Sullivan CS (2012) RNA virus microRNA that mimics a B-cell oncomiR. Proc Natl Acad Sci U S A 109: 3077-3082. 62. Tycowski KT, Guo YE, Lee N, Moss WN, Vallery TK, et al. (2015) Viral noncoding RNAs: more surprises. Genes Dev 29: 567-584. 63. Ding SW, Voinnet O (2007) Antiviral immunity directed by small RNAs. Cell 130: 413-426. 64. He L, Hannon GJ (2004) MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev Genet 5: 522-531. 65. Maillard PV, Ciaudo C, Marchais A, Li Y, Jay F, et al. (2013) Antiviral RNA interference in mammalian cells. Science 342: 235-238. 66. Landgraf P, Rusu M, Sheridan R, Sewer A, Iovino N, et al. (2007) A mammalian microRNA expression atlas based on small RNA library sequencing. Cell 129: 1401-1414. 67. Backes S, Shapiro JS, Sabin LR, Pham AM, Reyes I, et al. (2012) Degradation of host microRNAs by poxvirus poly(A) polymerase reveals terminal RNA methylation as a protective antiviral mechanism. Cell Host Microbe 12: 200- 210. 68. Haasnoot J, Berkhout B (2011) RNAi and cellular miRNAs in infections by mammalian viruses. Antiviral RNAi: Concepts, Methods, and Applications: 23- 41. 69. Lee SI, Murthy SC, Trimble JJ, Desrosiers RC, Steitz JA (1988) Four novel U RNAs are encoded by a herpesvirus. Cell 54: 599-607. 70. Wassarman DA, Lee SI, Steitz JA (1989) Nucleotide sequence of HSUR 5 RNA from herpesvirus saimiri. Nucleic Acids Res 17: 1258. 71. Golembe TJ, Yong J, Battle DJ, Feng W, Wan L, et al. (2005) Lymphotropic Herpesvirus saimiri uses the SMN complex to assemble Sm cores on its small RNAs. Mol Cell Biol 25: 602-611. 72. Cazalla D, Yario T, Steitz JA (2010) Down-regulation of a host microRNA by a Herpesvirus saimiri noncoding RNA. Science 328: 1563-1566. 73. Lagos-Quintana M, Rauhut R, Yalcin A, Meyer J, Lendeckel W, et al. (2002) Identification of tissue-specific microRNAs from mouse. Curr Biol 12: 735-739. 74. Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P (2005) Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science 309: 1577-1581. 75. Li Y, Masaki T, Yamane D, McGivern DR, Lemon SM (2013) Competing and noncompeting activities of miR-122 and the 5' exonuclease Xrn1 in regulation of hepatitis C virus replication. Proc Natl Acad Sci U S A 110: 1881-1886. 76. Ginsberg HS (1962) Identification and classification of adenoviruses. Virology 18: 312-319. 77. Rowe WP, Huebner RJ, Gilmore LK, Parrott RH, Ward TG (1953) Isolation of a cytopathogenic agent from human adenoids undergoing spontaneous degeneration in tissue culture. Proc Soc Exp Biol Med 84: 570-573. 78. SHENK T (1996) Adenoviridae: The viruses and their replication. ; B.N. Fields DMK, and P.M. Howley, editor. New York: Lippincott-Raven. 2111–2148 p. 79. Crawford-Miksza LK, Schnurr DP (1996) Adenovirus serotype evolution is driven by illegitimate recombination in the hypervariable regions of the hexon protein. Virology 224: 357-367. 80. Echavarria M (2008) Adenoviruses in Immunocompromised Hosts. Clinical Microbiology Reviews 21: 704-715. 81. Bergelson JM, Cunningham JA, Droguett G, Kurt-Jones EA, Krithivas A, et al. (1997) Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5. Science 275: 1320-1323.

46 82. Gaggar A, Shayakhmetov DM, Lieber A (2003) CD46 is a cellular receptor for group B adenoviruses. Nature Medicine 9: 1408-1412. 83. Wu E, Trauger SA, Pache L, Mullen TM, Von Seggern DJ, et al. (2004) Membrane cofactor protein is a receptor for adenoviruses associated with epidemic keratoconjunctivitis. Journal of Virology 78: 3897-3905. 84. Svensson U, Persson R (1984) Entry of Adenovirus-2 into Hela-Cells. Journal of Virology 51: 687-694. 85. Varga MJ, Weibull C, Everitt E (1991) Infectious Entry Pathway of Adenovirus Type-2. Journal of Virology 65: 6061-6070. 86. Wiethoff CM, Wodrich H, Gerace L, Nemerow GR (2005) Adenovirus protein VI mediates membrane disruption following capsid disassembly. J Virol 79: 1992-2000. 87. McConnell MJ, Imperiale MJ (2004) Biology of adenovirus and its use as a vector for gene therapy. Hum Gene Ther 15: 1022-1033. 88. Gattoni R, Schmitt P, Stevenin J (1988) In vitro splicing of adenovirus E1A transcripts: characterization of novel reactions and of multiple branch points abnormally far from the 3' splice site. Nucleic Acids Res 16: 2389-2409. 89. Berk AJ (1986) Adenovirus promoters and E1A transactivation. Annu Rev Genet 20: 45-79. 90. Braithwaite AW, Cheetham BF, Li P, Parish CR, Waldron-Stevens LK, et al. (1983) Adenovirus-induced alterations of the cell growth cycle: a requirement for expression of E1A but not of E1B. J Virol 45: 192-199. 91. Nakajima T, Masuda-Murata M, Hara E, Oda K (1987) Induction of cell cycle progression by adenovirus E1A gene 13S- and 12S-mRNA products in quiescent rat cells. Mol Cell Biol 7: 3846-3852. 92. Martin ME, Berk AJ (1998) Adenovirus E1B 55K represses p53 activation in vitro. J Virol 72: 3146-3154. 93. Sundararajan R, Cuconati A, Nelson D, White E (2001) Tumor necrosis factor- alpha induces Bax-Bak interaction and apoptosis, which is inhibited by adenovirus E1B 19K. J Biol Chem 276: 45120-45127. 94. de Jong RN, van der Vliet PC, Brenkman AB (2003) Adenovirus DNA replication: protein priming, jumping back and the role of the DNA binding protein DBP. Curr Top Microbiol Immunol 272: 187-211. 95. Burgert HG, Maryanski JL, Kvist S (1987) "E3/19K" protein of adenovirus type 2 inhibits lysis of cytolytic T lymphocytes by blocking cell-surface expression of histocompatibility class I antigens. Proc Natl Acad Sci U S A 84: 1356-1360. 96. Benedict CA, Norris PS, Prigozy TI, Bodmer JL, Mahr JA, et al. (2001) Three adenovirus E3 proteins cooperate to evade apoptosis by tumor necrosis factor- related apoptosis-inducing ligand receptor-1 and -2. J Biol Chem 276: 3270- 3278. 97. Shaw AR, Ziff EB (1980) Transcripts from the adenovirus-2 major late promoter yield a single early family of 3' coterminal mRNAs and five late families. Cell 22: 905-916. 98. Dolph PJ, Racaniello V, Villamarin A, Palladino F, Schneider RJ (1988) The adenovirus tripartite leader may eliminate the requirement for cap-binding protein complex during translation initiation. J Virol 62: 2059-2066. 99. Cuesta R, Xi Q, Schneider RJ (2000) Adenovirus-specific translation by displacement of kinase Mnk1 from cap-initiation complex eIF4F. EMBO J 19: 3465-3474. 100. Reich PR, Forget BG, Weissman SM (1966) RNA of low molecular weight in KB cells infected with adenovirus type 2. J Mol Biol 17: 428-439. 101. Mathews MB (1975) Genes for VA-RNA in adenovirus 2. Cell 6: 223-229.

47 102. Soderlund H, Pettersson U, Vennstrom B, Philipson L, Mathews MB (1976) A new species of virus-coded low molecular weight RNA from cells infected with adenovirus type 2. Cell 7: 585-593. 103. Bhat RA, Metz B, Thimmappaya B (1983) Organization of the noncontiguous promoter components of adenovirus VAI RNA gene is strikingly similar to that of eucaryotic tRNA genes. Mol Cell Biol 3: 1996-2005. 104. Schramm L, Hernandez N (2002) Recruitment of RNA polymerase III to its target promoters. Genes & Development 16: 2593-2620. 105. Railey JF, Wu GJ (1988) Organization of Multiple Regulatory Elements in the Control Region of the Adenovirus Type-2-Specific Varna1-Gene - Fine Mapping with Linker-Scanning Mutants. Molecular and Cellular Biology 8: 1147-1159. 106. Rohan RM, Ketner G (1987) A comprehensive collection of point mutations in the internal promoter of the adenoviral VAI gene. J Biol Chem 262: 8500-8507. 107. Snouwaert J, Bunick D, Hutchison C, Fowlkes DM (1987) Large numbers of random point and cluster mutations within the adenovirus VA I gene allow characterization of sequences required for efficient transcription. Nucleic Acids Res 15: 8293-8303. 108. Lassar AB, Martin PL, Roeder RG (1983) Transcription of class III genes: formation of preinitiation complexes. Science 222: 740-748. 109. Schramm L, Hernandez N (2002) Recruitment of RNA polymerase III to its target promoters. Genes Dev 16: 2593-2620. 110. Vennstrom B, Pettersson U, Philipson L (1978) Two initiation sites for adenovirus 5.5S RNA. Nucleic Acids Res 5: 195-204. 111. Thimmappaya B, Jones N, Shenk T (1979) A mutation which alters initiation of transcription by RNA polymerase III on the Ad5 chromosome. Cell 18: 947- 954. 112. Ma Y, Mathews MB (1996) Structure, function, and evolution of adenovirus- associated RNA: a phylogenetic approach. J Virol 70: 5083-5099. 113. Thimmappaya B, Weinberger C, Schneider RJ, Shenk T (1982) Adenovirus VAI RNA is required for efficient translation of viral mRNAs at late times after infection. Cell 31: 543-551. 114. Bhat RA, Thimmappaya B (1984) Adenovirus mutants with DNA sequence perturbations in the intragenic promoter of VAI RNA gene allow the enhanced transcription of VAII RNA gene in HeLa cells. Nucleic Acids Res 12: 7377- 7388. 115. Punga T, Kamel W, Akusjarvi G (2013) Old and new functions for the adenovirus virus-associated RNAs. Future Virology 8: 343-356. 116. Svensson C, Akusjarvi G (1984) Adenovirus VA RNAI: a positive regulator of mRNA translation. Mol Cell Biol 4: 736-742. 117. Reichel PA, Merrick WC, Siekierka J, Mathews MB (1985) Regulation of a protein synthesis initiation factor by adenovirus virus-associated RNA. Nature 313: 196-200. 118. Kitajewski J, Schneider RJ, Safer B, Munemitsu SM, Samuel CE, et al. (1986) Adenovirus VAI RNA antagonizes the antiviral action of interferon by preventing activation of the interferon-induced eIF-2 alpha kinase. Cell 45: 195- 200. 119. Mathews MB, Shenk T (1991) Adenovirus virus-associated RNA and translation control. J Virol 65: 5657-5662. 120. Katze MG, DeCorato D, Safer B, Galabru J, Hovanessian AG (1987) Adenovirus VAI RNA complexes with the 68 000 Mr protein kinase to regulate its autophosphorylation and activity. EMBO J 6: 689-697.

48 121. McKenna SA, Lindhout DA, Shimoike T, Aitken CE, Puglisi JD (2007) Viral dsRNA inhibitors prevent self-association and autophosphorylation of PKR. J Mol Biol 372: 103-113. 122. Maran A, Mathews MB (1988) Characterization of the double-stranded RNA implicated in the inhibition of protein synthesis in cells infected with a mutant adenovirus defective for VA RNA. Virology 164: 106-113. 123. Mellits KH, Mathews MB (1988) Effects of mutations in stem and loop regions on the structure and function of adenovirus VA RNAI. EMBO J 7: 2849-2859. 124. Sadler AJ, Williams BR (2008) Interferon-inducible antiviral effectors. Nat Rev Immunol 8: 559-568. 125. Desai SY, Patel RC, Sen GC, Malhotra P, Ghadge GD, et al. (1995) Activation of interferon-inducible 2'-5' oligoadenylate synthetase by adenoviral VAI RNA. J Biol Chem 270: 3454-3461. 126. Meng H, Deo S, Xiong S, Dzananovic E, Donald LJ, et al. (2012) Regulation of the interferon-inducible 2'-5'-oligoadenylate synthetases by adenovirus VA(I) RNA. J Mol Biol 422: 635-649. 127. Lu S, Cullen BR (2004) Adenovirus VA1 noncoding RNA can inhibit small interfering RNA and MicroRNA biogenesis. J Virol 78: 12868-12876. 128. Bennasser Y, Chable-Bessia C, Triboulet R, Gibbings D, Gwizdek C, et al. (2011) Competition for XPO5 binding between Dicer mRNA, pre-miRNA and viral RNA regulates human Dicer levels. Nat Struct Mol Biol 18: 323-327. 129. Xu N, Segerman B, Zhou X, Akusjarvi G (2007) Adenovirus virus-associated RNAII-derived small RNAs are efficiently incorporated into the rna-induced silencing complex and associate with polyribosomes. J Virol 81: 10540-10549. 130. Xu N, Gkountela S, Saeed K, Akusjarvi G (2009) The 5'-end heterogeneity of adenovirus virus-associated RNAI contributes to the asymmetric guide strand incorporation into the RNA-induced silencing complex. Nucleic Acids Res 37: 6950-6959. 131. Aparicio O, Carnero E, Abad X, Razquin N, Guruceaga E, et al. (2010) Adenovirus VA RNA-derived miRNAs target cellular genes involved in cell growth, gene expression and DNA repair. Nucleic Acids Res 38: 750-763. 132. Bellutti F, Kauer M, Kneidinger D, Lion T, Klein R (2015) Identification of RISC-Associated Adenoviral MicroRNAs, a Subset of Their Direct Targets, and Global Changes in the Targetome upon Lytic Adenovirus 5 Infection. Journal of virology 89: 1608-1627. 133. Zhao H, Chen M, Tellgren-Roth C, Pettersson U (2015) Fluctuating expression of microRNAs in adenovirus infected cells. Virology 478: 99-111. 134. Furuse Y, Ornelles DA, Cullen BR (2013) Persistently adenovirus-infected lymphoid cells express microRNAs derived from the viral VAI and especially VAII RNA. Virology 447: 140-145. 135. Alemany R, Balague C, Curiel DT (2000) Replicative adenoviruses for cancer therapy. Nat Biotechnol 18: 723-727. 136. Zhang Y, Huang W, Ornelles DA, Gooding LR (2010) Modeling adenovirus latency in human lymphocyte cell lines. J Virol 84: 8799-8810. 137. Zheng Y, Stamminger T, Hearing P (2016) E2F/Rb Family Proteins Mediate Interferon Induced Repression of Adenovirus Immediate Early Transcription to Promote Persistent Viral Infection. PLoS Pathog 12: e1005415.

49 Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1216 Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine”.)

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-283080 2016