Recombinaison expérimentale intra-et interespèce chez les Rhinovirus et Quantification d'arn de Rhinovirus par RT-PCR en temps réel.

Transcription

1 Thesis Recombinaison expérimentale intra-et interespèce chez les Rhinovirus et Quantification d'arn de Rhinovirus par RT-PCR en temps réel SCHIBLER, Manuel Abstract Les rhinovirus sont des petits virus à ARN positif appartenant au genre Enterovirus de la famille des Picornaviridae, caractérisée par une importante variabilité génétique acquise par mutation et recombinaison. La première partie de cette thèse est consacrée à l'étude de la recombinaison des rhinovirus en utilisant des génomes chimériques synthétisés in vitro ainsi que la recombinaison non réplicative, résultant de la co-transfection de génomes défectifs complémentaires. Nous avons ainsi observé que la région 5'non codante est interchangeable entre membres d'espèces differentes, alors que seule la recombinaison intraespèce semble viable dans la région codante. Les sites de recombinaison obtenus sont décrits. La seconde partie de ce travail consiste en une analyse critique des erreurs possibles liées à la quantification d'arn de rhinovirus dans des échantillons cliniques par RT-PCR en temps réel. Différentes expériences de validation montrent qu'une quantification relative est possible avec une marge d'erreur de plus ou moins 10%. Reference SCHIBLER, Manuel. Recombinaison expérimentale intra-et interespèce chez les Rhinovirus et Quantification d'arn de Rhinovirus par RT-PCR en temps réel. Thèse de doctorat : Univ. Genève, 2013, no. Sc. Méd. 13 URN : urn:nbn:ch:unige DOI : /archive-ouverte/unige:31100 Available at: Disclaimer: layout of this document may differ from the published version.

2 Section de médecine fondamentale Département de médecine génétique et de laboratoire Service de médecine de laboratoire Thèse préparée sous la direction du Docteur Caroline Tapparel et du Professeur Laurent Kaiser Recombinaison Expérimentale Intra- et Interespèce chez les Rhinovirus et Quantification d ARN de Rhinovirus par RT-PCR en Temps Réel Thèse présentée à la Faculté de Médecine de l'université de Genève pour obtenir le grade de Docteur en Sciences médicales «MD-PhD» par Manuel David SCHIBLER de Biberist (SO) Thèse n 13 Genève 2013

3

4

5 Aknowledgments I would like to sincerely thank the following people who directly or indirectly contributed to this thesis: Laurent Kaiser, for your trust, support, guidance and generosity throughout this long-lasting process Caroline Tapparel, for your precious expertise in the rhinovirus field, and for your patience and commitment in making an MD do biology Lara Turin and Sandra Van Belle, for your technical help Samuel Cordey, for your constant availability and willing to help, especially when experiments became tricky Yves Thomas and Pascal Cherpillod, for your advice and the nice atmosphere in the office Lorena Sacco and Michael Bel, for the nice breaks Sabine Yerly, for your analytical assistance, particularly during the rhinovirus RNA quantification project All other members of the Virology Laboratory, for the good times and nice working atmosphere Dominique Garcin and Laurent Roux, for your availability and all the discussions we have had concerning my projects Manel Essaidi, Carole Bampi and Geneviève Mottet-Osman for the fun we have had while you were teaching me the secrets of Western Blots Mylène Docquier, for your useful advices regarding RNA quantification by real-time RT- PCR Jérôme Pugin, Dominique Garcin, Francis Delpeyroux, Caroline Tapparel and Laurent Kaiser, for constituting such a great thesis Jury My wife Lamarana, my parents Monika and Ueli, my sister Muriel and my friends, for their precious presence during this important period of my life

6 UNIVERSITY OF GENEVA FACULTY OF MEDICINE Doctor Caroline Tapparel Professor Laurent Kaiser Experimental intra- and interspecies Rhinovirus Recombination and Rhinovirus RNA quantification by real-time RT-PCR THESIS Presented to the Faculty of Medicine of the University of Geneva for the MD-PhD Doctorate in Medical Sciences by Manuel David SCHIBLER from Biberist (Switzerland) Jury members: Prof Jérôme Pugin - Faculty of Medicine of Geneva and Geneva University Hospitals (President) Dr Caroline Tapparel - Faculty of Medicine of Geneva and Geneva University Hospitals (Thesis Director) Prof Laurent Kaiser - Faculty of Medicine of Geneva and Geneva University Hospitals (Thesis Co-Director) Dr Dominique Garcin - Faculty of Medicine of Geneva, University of Geneva (Local Expert) Prof Francis Delpeyroux - Pasteur Institute, France (External Expert) 1

7 Table of contents Introduction... 5 RHINOVIRUS CLASSIFICATION... 5 VIRION STRUCTURE AND GENOMIC ORGANIZATION... 7 RHINOVIRUS LIFE CYCLE Viral entry HRV translation and polyprotein processing Replication Autossembly and release MECHANISMS UNDERLYING HRV GENETIC VARIABILITY High mutation rate Recombination HRV VERSUS HEV PHENOTYPES in cell culture HRV TRANSMISSION, IMMUNE RESPONSE AND PATHOGENESIS Route of transmission and site of infection Immediate host defense mechanisms: innate immunity Cellular and humoral immune responses Pathogenesis CLINICAL MANIFESTATIONS RELATED TO HRV INFECTIONS Rhinosinusitis and otitis Lower respiratory tract infection Exacerbation of underlying airways diseases Asymptomatic patients EPIDEMIOLOGY RHINOVIRUS DETECTION

8 HRV isolation in cell culture HRV antigen detection HRV RNA detection ANTIVIRAL DRUGS AGAINST HRVs ANTI-HRV VACCINES Objectives EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION RHINOVIRUS RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS Materials and methods PLASMIDS AND CONSTRUCTS Construction of the chimeric HRV P1-2A genomes Construction of the deleted HRV parental genomes used in non replicative recombination experiments RNA EXTRACTION AND REVERSE TRANSCRIPTION IN VITRO TRANSCRIPTION AND TRANSFECTION CELL CULTURE VIRAL CULTURE IMMUNOFLUORESCENCE SEQUENCING AND MAPPING OF RECOMBINATION SITES Results EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION Chimeric HRV genomes generated by molecular cloning methods Non replicative recombination ability of artificially engineered defective viral genomes

9 RHINOVIRUS RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS: ARTICLE Discussion EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION Artificially engineered chimeric HRV genomes Experimental non replicative HRV RNA recombination HRV RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS CONCLUSIONS AND PROSPECTS References Appendix ARTICLE

10 Introduction RHINOVIRUS CLASSIFICATION Although taxonomic criteria are under constant debate and evolution, the International Committee on Taxonomy of Viruses (ICTV) developed a reference strain for virus classificiation ( Based on this system, the Picornaviridae family, belonging to the Picornavirales order, is currently divided into 12 genera: Enterovirus, Cardiovirus, Hepatovirus, Parechovirus, Erbovirus, Kobuvirus, Teschovirus, Sapelovirus, Senecavirus, Tremovirus and Avihepatovirus (Figure 1, and Recently, five new genera were proposed: "Aquamavirus", "Cosavirus", "Dicipivirus", "Megrivirus" and "Salivirus". Human pathogens are found among the Enterovirus, Cardiovirus, Hepatovirus, Parechovirus and Kobuvirus genera [1]. The Enterovirus genus, the largest of the family, is constituted of 10 human species: Human enterovirus A, Human enterovirus B, Human enterovirus C, Human enterovirus D, Simian enterovirus A, Bovine enterovirus, Porcine enterovirus B, Human rhinovirus A, Human rhinovirus B and Human rhinovirus C (Figure 1). According to the International ICTV, two members of the Enterovirus genus are considered to belong to the same species if they share more than 70 % amino acid (aa) identity in P1; if they share greater than 70 % aa identity in the non-structural proteins 2C and 3CD; if they share a limited range of host cell receptors; if they share a limited natural host range; if they have a genome base composition (GC) that varies by no more than 2.5 %; and if they share a significant degree of compatibility in proteolytic processing, replication, encapsidation, and genetic recombination (International 5

11 Committee on Taxonomy of Viruses, rate-official/1201.aspx). Each species is further divided in several types. Concerning the rhinovirus species, 100 Rhinovirus (HRV) serotypes were historically defined by crossneutralization properties in vitro [2, 3]. This classification obviously concerned only cultivable viruses. Later, a correlation between phylogenetic relationships of capsid proteins encoding sequences in the VP4/VP2 region or VP1 and serotypes has been established [4, 5] and HRVs are now classified by sequence similarities. Accordingly, the term serotype tends to be replaced by the terms genotype or simply type. A nucleotide sequence divergence in the VP1 gene higher than 12% for HRV-A and -B and than 13% for HRV-C separates two strains into different types inside the same species [6]. This classification system also allows type assignment for viruses that are not cultivable and thus cannot be submitted to serological characterization, such as HRV-Cs. It is noteworthy that the 5 UTR region, used as a PCR target to detect HRV, is not suited for reliable HRV typing, notably because HRV-A and HRV-Ca 5 UTR sequences cluster together [7] (see HRV recombination below). HRVs are currently divided into 153 types (77 HRV-A types, 25 HRV-B types and 51 HRV- C types) (Figure 1 and Nick Knowles, 2012, picornaviridae.com, This number will likely increase rapidly with the ongoing sequencing of previously undetected strains. 6

12 Figure 1. Picornavirus genera and enterovirus species The picornavirus genera are indicated in bold and the enterovirus species in normal characters. Black, grey, and white panels indicate genera and species with viruses infecting respectively animals without humans, animals including humans, and humans only. The proposed new genera or species are indicated in italics. The number of genotypes according to the host is indicated in brackets for each enterovirus species. SAFV, saffold Attachment Virus; EV, enterovirus; BEV, bovine enterovirus; PEV, porcine enterovirus; SEV, simian enterovirus; RV, rhinovirus; swi, swine; sim, simian; hum, human (Adapted from [8]). VIRION STRUCTURE AND GENOMIC ORGANIZATION The HRV genome is protected by a non-enveloped icosahedral capsid constituted of four structural proteins, VP1, VP2, VP3 and VP4. Most of the capsid s interaction with the environment resides in the VP1 protein. VP1, and to a lesser extent VP2 and VP3 proteins are the most exposed viral proteins and the targets of the immune response [9-11]. VP4 is localized on the inner surface of the capsid and interacts with genomic RNA [12]. 7

13 Each of the capsid protein assembles to form a protomer, and 60 protomers assemble to form the icosahedral capsid. Five VP1 tips join and form a five-fold symmetry axis (Figure 2). There are 12 such five-fold axis in an icosahedral capsid. VP2 and VP3 tips join and form a three-fold symmetry axis. There is a valley separating the VP1 surfaces from the VP2 and VP3 surfaces that surrounds each five-fold axis. In HRVs this valley is particularly deep and is referred to as the canyon [13]. The so-called hydrophobic pocket situated at the bottom of that canyon is involved in receptor binding and is an antiviral drug target [14, 15]. Although there is no sequence homology among the structural proteins, VP1, VP2 and VP3 all display an eight-stranded anti-parallel β barrel structure. These β strands are separated from each other by loops that protrude out from the viral capsid. These loops, along with the NH2 and COOH ends of the capsid proteins, are variable among the different picornaviruses and are responsible for their different receptor-binding and immunogenic features [16]. Figure 2. A. Schematic HRV icosahedral capsid diagram showing the subunit organization and canyon (shaded). Thick lines encircle five protomers of VP1-VP4. Two five-fold symmetry axis (black pentagons), two three-fold symmetry axis (black triangles), and three two-fold symmetry axis (black eye shaped symbols) are indicated (adapted from [16]). 8

14 The HRV genome consists of a single positive-sense RNA whose length is about bases. It can be subdivided into four parts: a 5 untranslated region (5 UTR); a single open reading frame (ORF); a short 3 untranslated region (3 UTR); and a poly(a) tail [17] (Figure 3A). The open reading frame codes for a polyprotein that is co- and post-translationally cleaved into four structural proteins (VP4, VP2, VP3 and VP1) and seven non-structural proteins (2A, 2B, 2C, 3A, 3B, 3C and 3D). The genomic RNA also harbours a stem-loop structure called the cis-acting replication element (cre), which is essential for replication initiation. In HRVs, this cre stem-loop structure is part of the ORF, and its location varies in the three HRV species. It lies within the 2A coding region for HRV-A, VP1 for HRV-B and VP2 for HRV-C (Figure 3B). In contrast, all HEV species have their cre situated in the 2C region [18]. A conserved AAAC sequence, found in all currently known picornavirus cre loop sequences (Figure 3B), is implicated in virion protein genome-linked (VPg) urydylation, a process essential to prime picornavirus replication. The Enterovirus 5 UTR folds into six structurally distinct domains that can be divided into two functional units: a 5 cloverleaf structure (CL) essential for replication (domain I) and a type I internal ribosome entry site (IRES) (domains II-VI) necessary for the cap-independent translation of the polyprotein ORF [19] (Figure 3C). The CL contains four structural domains: stem A, stem-loop B, stem-loop C and stem-loop D. Stem-loop B is a known binding site for poly(rc)-binding protein 2 (PCBP2), and stem-loop D interacts with the viral 3C and 3CD proteases. The functions of stem A and stem-loop C are currently unknown. The CL, cellular PCBP2 and viral 3C and 3CD are all implicated in the regulation of picornavirus translation and replication. Two different models describe putative mechanisms for the switch from translation to replication regarding the poliovirus life cycle. In the first model, PCBP2 binding to the CL stem-loop B stimulates the PV polyprotein translation, and once a sufficient amount of viral 3CD is present, it binds to stem-loop D to repress translation. The positive strand PV 9

15 RNA thereby becomes available for minus strand RNA synthesis [20]. Alternatively, according to the second and more recent model, PCBP2 binds to the IRES stem-loop IV and thereby stimulates translation. 3C and 3CD proteinases resulting from viral polyprotein translation then cleave PCBP2. This cleavage releases a specific domain, KH3. The remaining PCBP2 KH3 protein, whose affinity for stem-loop IV but not for the CL stem-loop B is lost, is released from the IRES, and translation is no longer stimulated, leaving the viral genome accessible for replication [21]. Although replication regulation has been mainly attributed to the CL, sequences located in the 3 region of the poliovirus (PV) IRES may also be implicated [22]. The viral protein 3B (or VPg), is covalently bound to the 5 extremity of the viral genome. Its role in HRV RNA synthesis is discussed in the Replication section. The HRV 3 UTR, whose length ranges from 40 to 60 nuctleotides (nt), includes a 13 to 16 nt long stem, ending with the poly(a) tail. The function of this stem is not elucidated, but it may be implicated in protein binding or RNA:RNA interactions with other parts of the genome during the replication process. It has been shown that the 3 UTR is not essential for poliovirus RNA replication, but its presence greatly enhances its efficiency [23]. 10

16 A. B. C. IV. VI. V. C D I. II III. B A Figure 3. A. HRV genomic organization (adapted from [17]). B. Sequence and secondary structure of an HRV- A, an HRV-B and an HRV-C cre stem-loop. The conserved AAAC key-sequence is shown in bold (adapted from [24]). C. Enterovirus 5 UTR structure depicting the 6 characteristic domains (I-VI). The subdomains of the cloverleaf structure are also shown (A-D) (adapted from [25]). 11

17 RHINOVIRUS LIFE CYCLE Viral entry Rhinoviruses use different receptors. Most HRV-A and all HRV-B bind to intracellular adhesion molecule 1 (ICAM-1) via the above mentioned canyon region of the capsid. In contrast, 11 HRV-A members [26] use the low density lipoprotein receptor (LDLR), which binds to the five-fold axis. Finally, the HRV-C receptor(s) is(are) currently unknown, and seem(s) to be different from ICAM-I and LDLR, based on bioinformatic comparisons and inhibition assays [27, 28]. In silico studies suggest that an ICAM-I-like molecule may serve as an HRV-C receptor [29]. HRV entry has been studied intensely for HRV-A2 and HRV-B14. However, the precise mechanisms concerning viral particle conformational changes, internalization, uncoating and RNA release into the cytosol remain largely unknown (reviewed in [30]). Binding of HRV to its specific receptor triggers conformational changes in the capsid. This initial conformational change is necessary for receptor-mediated endocytosis and further structural changes induced by the low-ph endosomal environment that are responsible for uncoating and RNA release into the cytoplasm [30](Figure 4). This likely involves endosomal membrane rupture or pore formation by viral capsid proteins. The fact that empty HRV-14 capsids have been highlighted in the cell cytosol favors the endosomal rupture model [30]. It is hypothesized that the viral RNA might be released into the cytosol concomitantly to this process. 12

18 Figure 4. Model for HRV entry and uncoating. The virus binds to its receptor (ICAM-1 or LDLR) at the plasma membrane (1.), is uptaken by various endocytic pathways depending on the cell type (2.) and is delivered to early endosomes (3.). Viral capsid conformational changes occur following receptor binding and exposure to low ph in late endosomes (4.), resulting in the release of viral capsids and HRV RNA release into the cytoplasm (5.). Receptors present in the early endosomes are recycled to the cell surface (6.). HRV translation and polyprotein processing The RNA released in the cytoplasm is of positive polarity and thus directly translated by host cell ribosomes. For translation to occur, VPg is cleaved from the 5 end of the viral genome by 13

19 a cellular enzyme named VPg-unlinkase. This enzyme has been proposed to serve as a marker to distinguish viral RNAs used for translation (devoid of VPg) from those implicated in replication (VPg-linked) [31]. Picornaviruses ORF translation is cap-independent and mediated via the IRES. The eukaryotic initiation factors eif4g, eif4a, and eif4b are recruited to the 5 UTR domain V of the IRES. This ribonucleoprotein (RNP) complex in turn recruits the 40S ribosomal subunit associated with eif3 and eif2-gtp-met-trna. The 40S subunit then scans the 3 region of the 5 UTR until it reaches the authentic start codon where the complete ribosome is assembled [32]. In addition to bypass cap-dependant translation, HRVs efficiently inhibit host cell translation. As shown for poliovirus, HRV 2A protease cleaves the eif4g factor (formerly p220), whose integrity is essential for cap-dependant translation [33]. Various cellular proteins called IRES trans-acting factors (ITAFs) are also implicated in HRV translation regulation [34]. Translation results in the synthesis of a viral polyprotein precursor of about 2000 aa which is co- and post-translationally cleaved into smaller protein products. The first cleavage occurs between the P1 and P2 region and is mediated by the viral 2A proteinase. The second cleavage is made by the 3C proteinase at the P2-P3 junction (Figure 5). All subsequent cleavages are carried out by the 3C proteinase, except for the cleavages between VP0 and VP3 and between VP3 and VP1 that are performed by the 3CD precursor proteinase [35]. VP0 cleavage into VP4 and VP2 occurs later in the viral life cycle, during progeny virions maturation, and is probably mediated by an autocatalytic process [36]. The various cleavage products and their functions are shown in Figure 5. Of note, some protein precursors, such as 2BC, 3AB and 3CD, display specific functions in the infected cell (Figure 5). As shown in Table 1, some polyprotein cleavage sites vary among the different HRV and HEV species. This may limit the interspecies HRV recombination possibilities, as discussed 14

20 later. The polyprotein cleavage sites of 7 HRV-A genomes, 3 HRV-B genomes, 3 HRV-C genomes, 1 HEV-A genome, 1 HEV-B genome, 2 HEV-C genomes, and 2 HEV-D genomes are shown in Table 1. As the number of HRV and HEV genomes used in this analysis is limited, the existence of additional cleavage sites cannot be excluded. Figure 5. HRV and HEV polyprotein processing cascade and mature protein functions. The primary and secondary polyprotein cleavages are depicted. The viral proteinases are highlighted in dark blue. The known functions of the various cleavage products are described (adapted from [37]). 15

21 Table 1. Polyprotein cleavage sequence according to the three HRV species and the four HEV species. Polyprotein cleavage site Protease HRV-A (HRV-A1, -A2, -A16, -A39, - A81, - A89, -A100) HRV-B (HRV-B14, -B84, - B42) HRV-C (HRV- C11, -C6, -C28) HEV-A (HEV-A71) HEV-B (HEV-B69) HEV-C (PV-1, CV- A20) HEV-D (HEV-D68, -D94) VP4-VP2 - Q-S N-S M-S K-S S-P N-S L-S VP2-VP3 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G VP3-VP1 3C Q-N E-G Q-N Q-G Q-N Q-G Q-S Q-G D-L Q-L VP1-2A 2A V-G Y-G A-G L-G H-G Y-G T-G A-G L-G V-G 2A-2B 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G 2B-2C 3C Q-S Q-A Q-S Q-S Q-N Q-G Q-G E-S Q-S Q-G Q-S 2C-3A 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G Q-Y 3A-3B 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G 3B-3C 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G 3C-3D 3C Q-G Q-G Q-G Q-G Q-G Q-G Q-G F-E Replication Picornavirus replication is a two-step process taking place in membrane-associated replication complexes. The first step consists of the synthesis of a negative-sense full length RNA that is complementary to the viral positive sense genome and the second step implies the synthesis of new positive sense viral genomes from the negative sense RNA templates. As for translation initiation, the precise mechanisms implicated in HRV replication are not fully understood and most of the data derive from research involving poliovirus. 16

22 To initiate negative sense RNA synthesis, the picornavirus 3D RNA-dependant-RNApolymerase (RdRp) requires a primer. This primer, VPg-pU-pU, is constituted of the viral VPg protein, to which two uridine residues are covalently linked by the 3D polymerase by a process called VPg-uridylylation. This process is orchestrated by the cre, which serves as template via a specific and conserved AAAC sequence (Figure 3B) to which 3CD binds and thereby stimulates 3D-mediated VPg uridylylation. VPg-pU-pU remains linked to 3D, and acts as a primer, enabling negative sense RNA synthesis initiation (Figure 6) at the 3 end of the viral poly(a) tail [24]. Figure 6. Cre-dependent VPg uridylylation. The scheme depicts VPg uridylylation by 3D on the cre AAAC template, with 3CD being bound to the cre and promoting the process (adapted from [24]). In parallel to VPg uridylylation, according to a current model, the 3AB precursor protein associates with cellular membrane elements via its hydrophobic domain, while 3CD precursor protein binds to the D domain of the CL (see above) and PCBP2 binds to the B domain of the CL. The complex formed by the viral positive-strand RNA and these two proteins is recruited to cellular membrane elements via an interaction between membrane-associated 3AB and PCBP2. A cellular protein, poly(a)-binding protein (PABP), in turn binds to the viral genome 17

23 poly(a) tail, and interacts with 3CD and PCBP2 which are bound to the CL, thereby circularizing the viral RNA. This process is required for negative sense RNA synthesis initiation [38]. In addition to the 5 UTR and the cre, cis-active RNA structures implicated in replication are included in the 3 UTR. Although not essential for poliovirus replication, the 3 UTR has been shown to greatly increase replication efficiency [23, 39]. The poly(a) tail is required for replication and its length influences replication efficiency [40]. The double-stranded RNA helix constituted of a positive strand RNA and the complementary negative strand RNA is called the replicative form (RF) [41]. After synthesis of the minus strand RNA, new positive strand RNAs can be synthesized. In fact, 40 to 70 progeny positive strand RNAs derive from a single negative strand RNA [42]. Positive-strand synthesis is also initiated via VPg-pU-pU priming. In this instance, the two uridine residues hybridize to the two adenosine residues situated at the 3 end of the negative strand RNA. Partial denaturation of the RF is needed for this process to happen. How the separation of the two RNA strands of the RF occurs is not yet known [41, 43]. The newly synthesized positive RNA strands are either used as progeny genomes encapsidated into new viral particles or as templates for further translation. To date, it is still not elucidated whether cre-dependant VPg uridylylation is implicated in positive-strand RNA synthesis, in negative-strand RNA synthesis, or in both processes. Some researchers attribute a role of the cre motif only for positive-strand RNA synthesis [44, 45], whereas others have found evidence suggesting that cre-dependant VPg uridylylation is only involved in negative-strand RNA synthesis [46, 47]. As mentioned above, picornaviruses replication does not occur randomly in the infected cell cytoplasm. Like all known plus-strand RNA viruses, picornaviruses replicate in close 18

24 relationship to intracellular membranes. In the case of picornaviruses, these membranes seem to be derived from the endoplasmic reticulum-golgi apparatus network [48] and autophagosomes [49], and are rearranged to form vesicles, by the viral protein 2C [50]. The HRV 3A protein also seems to be implicated in membranous rearrangements [51]. The association of picornavirus replication to membranous vesicles represents a strategy providing several advantages: it enables the concentration of viral components to a defined compartment, it may promote the spatial organization of replication complexes, and the membranes may protect the viral replication machinery from the cellular innate immunity sensors [48]. Autossembly and release. Little is known about rhinovirus assembly and release processes. Rhinovirus assembly supposedly occurs in a fashion similar to that happening in polioviruses [52]. The first step is the assembly of VP1, VP3 and VP0 precursor into a 5S protomer, which is thought to form co-translationally. Five protomers then assemble into a 14S pentamer, and 12 pentamers are linked together to form the icosahedral empty 80S capsid. Whether this capsid forms around the viral RNA or the RNA enters pre-assembled capsids is still not elucidated. The final step in the maturation of the viral particle is the autocatalytic cleavage of VP0 into VP4 and VP2 [53]. Release of viral particles, at least concerning poliovirus, is classically thought to occur via cell lysis induced by the viral infection [54]. However, a more recent model suggests that vesicles derived from autophagosomes and induced by replication of poliovirus and HRVs might trap mature viral particles. These could ultimately be delivered into the extracellular medium by membrane fusion, thus avoiding cell lysis [49]. This would be in agreement with 19

25 the observation of intact respiratory epithelium during the course of in vivo HRV infections [55]. Figure 7 summarizes the key steps involved in the HRV life cycle. 20

26 Figure 7. Summary of the HRV life cycle (adapted from [54]). The virion recognizes its receptoir on the host cell s surface (1). The interaction between the viral capsid and the receptor triggers viral entry by endocytosis (2). Once the viral genome is released into the cytoplasm, it is directly translated by cellular ribosomes, in a CAP-independent manner (3). The viral polyprotein is co-translationnaly and autocatalytically processed into individual viral structural and non structural proteins (4). Once present in sufficient amounts, the viral 3D RNA polymerase initiates negative-strand RNA (5) and then positive-strand RNA (6). The latter is either used for further translation (7) or as a viral genome packed in newly assembled capsids during morphogenesis (8), leading to progeny virions which are released from the infected cell (9). 21

27 MECHANISMS UNDERLYING HRV GENETIC VARIABILITY As discussed above, there are more than 150 HRV genotypes ( [6]. The main explanations for such an important genetic variability are the high error rate of the viral RNA-dependant RNA polymerase and recombination [56]. High mutation rate The RNA-dependant RNA polymerase is characterized by a high mutation rate, estimated around 10-4 mutations per nucleotide. This represents almost one mutation per HRV genome per replication cycle [56]. During an infection, the viral population is thus not genetically homogeneous, but each genome differs by one or several point mutation(s). Such a viral population harbouring a cloud of related but slightly different genomes is called a quasispecies (Figure 8). The advantage of such a heterogeneous genetic repertoire is that upon environmental changes, one or several genomes may already be suited to the new environmental conditions. This or these will be selected and further mutate to recreate a diverse population. 22

28 Figure 8. Generation of a viral quasispecies. This scheme illustrates in a simplified manner how a quasispecies arises starting from a single viral genome. In these trees, each branch links together viral genomes differing by one point mutation, and each circle represents a replication cycle (Adapted from [57]). Recombination Viral recombination involves the exchange of genomic fragments between two different viruses. These two viruses need to be sufficiently similar to be compatible and to generate functional recombinant progeny. This classically occurs between two viruses belonging to the same species. Intraspecies recombination events have been extensively described regarding HEVs and are considered as an evolutive driving force for this virus group [58, 59]. The HEV recombination breakpoints mostly map around the 5 (VP4) and 3 ends (VP1-2AB junction) of the P1 region, while they are almost absent in the VP2-VP3-VP1 capsid region (Figure 9A)[58, 60]. A frequently observed HEV intraspecies recombination phenomenon concerns circulating vaccine-derived poliovirus (cvdpv) strains. These cvdpvs, whose pathogenicity can be 23

29 similar to that of wildtype PV, result from recombination between attenuated oral polioviruses and co-circulating non-poliovirus HEV-C members [61-63]. Some recent natural interspecies recombination events have also been described among circulating HEVs [64-66]. The recombination sites in these instances have been mapped to the 5 UTR and the 3D region (Figure 9B). Recombination events seem to occur less frequently among circulating HRVs. The recombination breakpoints identified such circulating HRV recombinants are situated at the 3 end of the 5 UTR and at the 5 end of the 3C gene (Figure 9C) [67]. Whether the difference in recombination frequency between the two virus groups is related to the type and site of infection, the frequency of co-infection or genomic features remains an open question. Phylogenetic studies indicate that interspecies HRV recombination occurred in the past. For instance, recombination between the 5 UTR of HRV-A and the polyprotein of HRV-C was proposed as the mechanism at the origin of the HRV-Ca subgroup that harbours HRV-A-like 5 UTR sequences [68, 69]. The remainder HRV-C strains, called HRV-Cc, exhibit 5 UTR sequences divergent from those of HRV-A, HRV-B and HRV-Ca members. Three putative interspecies recombination breakpoints in the 5 UTR have been mapped for HRV-Ca strains around position 481, 565, in the polypyrimidine tract, and 523, within stem-loop 5 of the IRES (Figure 9D)[69]. In the majority of the sequences analyzed, recombination presumably occurred in either one of the last two recombination hotspots, which are located in highly conserved sequence stretches. These two particular locations may therefore represent preferred sites for other interspecies 5 UTR recombination within the Enterovirus genus. Furthermore, some HRV-C strains harbour short HRV-A sequences in their 2A region (Figure 9D) [68, 69]. Analysis of the full-length sequences of all known HRV types in 2009 suggests that some HRV types resulted from ancient intraspecies recombination [70]. The majority of 24

30 the recombination sites revealed by this analysis are situated in the 5 UTR and the adjacent P1 region (Figure 9E). A. B. C. D. E. Figure 9. Major genomic regions involved in HEV and HRV recombination (in red). HEV intraspecies recombination among circulating strains (A). HEV interspecies recombination sites among circulating strains (B). HRV intraspecies recombination sites among circulating strains (C). Ancient HRV interspecies recombination between HRV-A and HRV-C species (D). Ancient HRV intraspecies recombination resulting in new HRV types (E). Finally, based on full genome phylogenetic analysis, it was proposed that ancient recombination events between HRV-A and HEV members gave rise to the HRV-B species (Figure 10) [71]. 25

31 Figure 10. The HRV-B species might have arisen from recombination between HRV-A and HEV members. The whole-polyprotein, maximum likelihood phylogenetic tree shows a closer relation between HRV-B and HEV than between HRV-A and HRV-B. The percentage of bootstraps (out of 1000) supporting corresponding clades is indicated. The sequence of simian picornavirus 1 (SV-2) was used as an outgroup. The branch lengths are measured in substitutions per site (adapted from [71]). 26

32 However, based on sequence homology, all the above proposed natural interspecies recombination events likely occurred between ancestors of the current HRV circulating strains. Two RNA virus recombination models have been proposed. The first is referred to as the copy-choice or template switch model (Figure 11A). The viral RNA polymerase synthesizes an RNA fragment complementary to the positive sense RNA template originating from one of the two viruses co-infecting the same cell, before being released from that template and resuming RNA synthesis on a neighbour genome originating from the other infecting virus [72]. Gmyl et al unraveled an alternative poliovirus recombination mechanism in vitro, called non replicative RNA recombination (Figure 11B), in which two different viral genomes are cleaved and joined together, resulting in a chimeric genome [73]. In their study, deleted poliovirus genomes were designed to promote non replicative recombination in the spacer region of the poliovirus RNA, situated between the IRES and the coding sequence and known to tolerate sequence variations without altering viability. Poliovirus genomes consisting of intact CL and IRES elements but lacking the polyprotein and 3 UTR sequences were co-transfected with genomes harbouring intact polyprotein and 3 UTR sequences but lacking an essential cis-acting element of the 5 UTR. Rescued polioviruses were sequenced, allowing recombination sites mapping. The vast majority of them were found in the spacer region. The exact molecular reactions underlying this type of recombination are not understood. While both mechanisms may be implicated in recombination events, it is currently not possible to determine which one is predominant, although it is generally believed that the copy-choice model is responsible for natural recombination. 27

33 A. B. Figure 11. RNA viruses recombination models. Template switch model (A, adapted from [74]), and the non replicative RNA recombination model (B, adapted from [75]. HRV VERSUS HEV PHENOTYPES IN CELL CULTURE In this section, in vitro HRV and HEV different phenotypic characteristics are described. Two characteristics have been used in the past to differentiate HRVs and HEVs. The first is the optimal growth temperature. While enteroviruses replicate efficiently at 37 C, rhinoviruses were found to have an optimal growth temperature of 33 C, the temperature of the upper respiratory tract. This rule is not absolute however, as some respiratory enteroviruses such as EV-D68 multiply the best at 33 C [76], and several HRVs replicate well at 37 C [77]. The second characteristic that differentiates both virus groups is acid tolerance. 28

34 Most HEVs are resistant to acidic ph values of 3 and even lower [54], which is not surprising since they need to survive the acidic environment of the stomach, before reaching their replication site in the intestine. In contrast, HRVs are labile at ph below 6 [54], and completely inactivated at ph 3 [78]. Again, some enteroviruses, such as EV-D68 are acid labile [76]. HRVs and HEVs exhibit very different cell culture tropism. HRVs are mainly cultivated in human diploid fibroblast cell lines such as embryonic lung fibroblast cell lines WI-38 and MRC-5, and HeLa Ohio cells overexpressing ICAM-1 [79, 80]. Of note, the recently discovered HRV-Cs are not cultivable in standard cell lines, and as mentioned earlier, their receptor(s) is(are) not yet identified. HEVs grow in a wide variety of human and non human cell lines [81]. However a given cell line is susceptible to some HEVs and not to others, depending on receptor usage [82]. Common cell lines used for HEV isolation include human rhabdomyosarcoma cells (RD), human embryonic lung cells (MRC-5), laryngeal cancer cells (Hep-2), human lung cancer cells (A549) and African green monkey kidney cells (Vero), among many others [81]. HRV TRANSMISSION, IMMUNE RESPONSE AND PATHOGENESIS Route of transmission and site of infection Experimental data suggest that HRVs are transmitted mainly by hand carriage, and infection occurs mostly via eye or nose, more rarely by inoculation into the mouth [83]. Aerosols appear to be an alternative transmission route [84, 85]. HRV particles deposited on the eye and transported to the nasal cavity via the lachrymal duct, and those deposited into the nose 29

35 are carried via the mucociliary system to the nasopharynx, the major HRV replication site [86]. Immediate host defense mechanisms: innate immunity Once the HRV genome is released in an infected cell, it is recognized by Toll-like receptors (TLR) and by a cytoplasmic pattern recognition receptor (PRR). The HRV capsid is detected by TLR2 on the epithelial surface, and HRV ssrna and dsrna are recognized by TLR3, TLR7 and TLR8 [87]. The cytoplasmic PRR involved in HRV genome detection is melanoma differentiation-associated gene 5 (MDA-5), which recognizes long double-stranded RNA [88, 89]. After having recognized picornaviral double-stranded RNA (i.e. the RF), MDA-5 activates interferon-β promoter simulator-1 (IPS-1) (also called Cardif, MAVS and VISA). IPS-1 in turn activates the kinases TBK-1 and IKKε, which phosphorylate interferon regulatory factors 3 and 7 (IRF-3 and IRF-7). These transcription factors induce type I interferons (IFN) (IFN-β and IFN-α) gene expression [90] (Figure 12). 30

36 Figure 12. Cytosolic recognition of various RNA viruses. The signaling pathway triggered by picornavirus double-stranded RNA detection is shown on the left (adapted from [91]). Type I IFNs are secreted and act in an autocrine and paracrine manner by binding cell surface type I interferon receptor, which triggers an intracellular signaling cascade implicating the Jak/STAT pathway [92] (Figure 13). This results in the expression of more than several hundred interferon stimulated genes (ISGs), inducing an antiviral state that inhibits viral replication [93]. 31

37 Figure 13. Type I interferon signaling via the Jak/STAT pathway (Adapted from [94]). However, this cellular defense mechanism, from viral detection to the establishment of the antiviral state, is the target of several viral proteins. For example, regarding viral detection by the host cell, MDA-5 is cleaved in a proteasome- and caspase-dependent manner in poliovirus infected cells [95], and IPS-1 is cleaved by HRV 2A and 3C proteases [96]. Furthermore, the antiviral state induced by type I IFNs can be reversed, at least to some point, as Enterovirus members (PV-1, PV-2, PV-3, EV-70, and HRV-A16) were shown to replicate in IFN-α treated cells. This ability was linked to the viral 2A protease [97]. Thus, a balance seems to exist between the host defense mechanisms that prevent massive tissue damage due to uncontrolled viral replication, and viral countermeasures that ensure sufficient HRV multiplication to allow viral transmission and survival. 32

38 Cellular and humoral immune responses The innate immune response triggered by rhinovirus infection, is implicated in the establishment of the adaptive cellular and humoral immune response. T-cells RANTES and IFN- -inducible protein (IP)-10 secreted by HRV-infected epithelial cells recruit T-cells to the infected airways [98]. It has been shown that a given CD4+ cell clone can be activated by exposure to different HRV serotypes, indicating the existence of shared epitopes [99]. Thus, memory CD4+ cells could be activated upon an HRV infection, and be implicated in a subsequent infection involving a serotype exhibiting common epitopes. The role of CD8+ T-cells in the immune response against HRVs is not yet defined. B-cells Anti-HRV antibodies are usually detected one to two weeks after inoculation, and the maximal titer is achieved five weeks after infection, with elevated titers persisting for at least one year [100]. Since the duration of an HRV infection is of a few days, with viral loads in nasal washes peaking at 48 to 72 hours post inoculation [101], the role of these antibodies is to prevent a subsequent infection involving the same HRV type rather than clearing the virus that triggers their synthesis [102, 103]. HRV-neutralizing antibodies protecting against another infection with the same HRV type are probably mucosal secretory IgAs [104]. However, as the number of different HRV types exceeds 150, a human individual can experience numerous HRV infections during his life. 33

39 Pathogenesis In addition to induce type 1 IFNs, viral detection by the host cell also results in the production of pro-inflammatory cytokines, via NF- B activation [105]. Several inflammatory mediators such as kinins, leucotriens, histamine, interleukins 1, 6 and 8, Tumor Necrosis Factor alpha (TNF-α) and Regulated upon Activation, Normal T-cell Expressed and Secreted (RANTES), were found to be increased in nasal secretions in the course of an HRV infection [106]. The inflammatory reaction induced by these molecules leads to vasodilatation and increased capillary permeability, causing rhinorrhea and nasal obstruction. Tissue damage resulting from HRV replication is almost inexistent, and the integrity of the respiratory epithelium remains intact upon an HRV infection [55]. This suggests that the pathogenic effects are mostly related to the innate immune response triggered by the infection. This hypothesis is supported by the correlation between interleukin-8 (IL-8) levels and symptoms following HRV infection [107]. CLINICAL MANIFESTATIONS RELATED TO HRV INFECTIONS HRVs are the most frequent agent of the common cold, an infection restricted to the upper respiratory tract. The major symptoms include nasal obstruction and discharge. Sneezing, sore throat and cough are often associated. Although this illness is benign and self-limited, it generates important costs due to medical seeking and absenteeism, and is associated with inappropriate antibiotics use. Furthermore, HRV can cause more severe clinical manifestations in selected populations. Of note, it is not yet clear whether a given HRV genotype or species is associated with a particular clinical presentation [8, ]. Results from a recent publication concerning HRV infections in infants suggest that HRV-A and 34

40 HRV-C species are associated with infections causing similar clinical manifestations that are more severe than those caused by HRV-B strains [112]. Rhinosinusitis and otitis HRV infections are not exclusively restricted to the nasal cavity and the pharynx, but also involve the para-nasal sinuses and are often associated with rhinosinusitis [106, 113]. In children, HRV infection frequently provokes a Eustachian tube dysfunction, which can lead to otitis media [114, 115]. Lower respiratory tract infection HRVs have long been considered unable to infect the lower respiratory tract. This belief was based on observations indicating that the optimal HRV growth temperature was 33 C [116]. However, these experiments involved only one HRV serotype. It has been shown since then that the higher temperature found in the lower respiratory tract is not necessarily restrictive for HRV replication [77], and that these viruses do indeed infect this site [ ]. Furthermore, HRVs have been identified as major pathogens of viral pneumonia among children, representing the most frequent virus group recovered from children with viral pneumonia in one study [123]. Exacerbation of underlying airways diseases The use of PCR enabled to establish an association between rhinovirus infections and exacerbation of chronic lung conditions, such as chronic obstructive pulmonary disease (COPD) [124], asthma [105, 125, 126] and cystic fibrosis [ ]. 35

41 It has been demonstrated that primary bronchial epithelial cells from asthmatic patients produced less IFN-β when compared to normal individuals [130]. It is therefore believed that viral clearance is decreased in these patients, resulting in increased inflammation and facilitated viral spread to the lower airways, explaining HRV-induced asthmatic exacerbations [98]. Asymptomatic patients Asymptomatic HRV infections have been documented for all age categories [ ] but seem to be more frequent among children [134]. However, asymptomatic chronic HRV shedding has not been clearly reported. Prolonged HRV shedding (8 months and longer) has only been documented among immunosuppressed patients [8, 121]. EPIDEMIOLOGY HRVs are circulating throughout the year in the human population, but HRV detection rates peak in autumn and spring [135]. However, HRV infections acquired during winter months showed to be more severe than those occurring during spring and summer, in an infant population [112]. They are represented worldwide [136], and no particular geographic distribution of the various HRV types has been documented. Various HRV types circulate concomitantly, without following a predictable scheme [135]. 36

42 RHINOVIRUS DETECTION The clinical specimens currently used for HRV detection are nasopharyngeal swabs (NPS), nasopharyngeal aspirates (NPA), nasal washes (NW), tracheal aspirates (TA), bronchial aspirates (BA) and bronchoalveolar lavages (BAL) [137] and the principal diagnostic means are virus isolation in cell culture and molecular assays. Electron microscopy is not used for routine diagnostics, due to the need for specialized equipment, the high costs linked to the technique, the slowness of the procedure and its reduced sensitivity compared to molecular diagnostic tools. Serology is neither used given the large number of different serotypes. HRV isolation in cell culture Assessment of viral multiplication in cell cultures via the observation of a cytopathic effect was the classical laboratory diagnostic method for HRV detection. Cell lines initially used for HRV isolation include primary human embryo kidney epithelial cells, human embryo kidney epithelial cells, and human embryo lung fibroblasts (HEL) [80]. These were replaced by semicontinuous strains of HEL cells, such as the WI-38 strain, and selected strains of HeLa cells overexpressing ICAM-1, the major HRV group receptor [79, 80]. An advantage of virus isolation is virus enrichment for phenotypic characterization. Furthermore, this method can rescue divergent strains that might escape molecular diagnostics due to sequence mismatches, provided that the virus is cultivable. However, this laboratory technique is time-consuming, requires expertise and its sensitivity is limited, as some HRV strains cannot be propagated on standard cell lines, and samples with low viral loads or poorly conserved samples will likely be culture-negative. 37

43 HRV antigen detection The large number of HRV serotypes renders HRV antigen detection difficult. This method is thus not used for HRV diagnosis in many virology laboratories. HRV RNA detection The major breakthrough in HRV diagnostics certainly regards the development of molecular techniques involving HRV RNA detection in clinical samples. The first HRV classic reversetranscription, polymerase chain reaction (PCR) assays were developed in the 1980s and targeted conserved sequences stretches in the 5 UTR of HRV genomes [ ]. This technique is far more sensitive than cell-culture, allowing the detection of small amounts of HRV RNA. However, conventional PCR remains time-consuming and tedious. The advent of real-time RT-PCR applied to HRV RNA detection further increased sensitivity and specificity, in addition to significantly reduce the time of the procedure [141, 142]. Diagnostic HRV real-time RT-PCR assay are designed to detect sequence stretches in the 5 UTR that are conserved among HRVs. With the ongoing discovery of new HRV strains and even previously unknown HRV species, the design of new real-time RT-PCR assays regularly needs to be adapted in order to detect as much different HRV genomes as possible [143]. However, this implies the use of primers and/or probes with degenerate positions that will inevitably favor the amplification of certain HRV genomes and disfavor the detection of others. One major advantage of real-time RT-PCR over cell culture is the possibility to detect HRV strains or species unable to be propagated in standard cell lines, as exemplified by the recent discovery of the HRV-C species. 38

44 Attempts are currently made to apply real-time RT-PCR to HRV RNA load quantification in respiratory specimens [144]. Such data are of clinical interest, as they allow studying the possibility of a correlation between viral load and disease severity. Furthermore, HRV RNA quantification could be used to monitor the course of chronic HRV infections in immunocompromised patients such as lung transplant recipients, and to explore possible associations between HRV load and graft rejection in these patients. In addition to real-time RT-PCR, other molecular techniques used to detect HRV RNA have recently been developed: in situ hybridization, which allows to highlight the presence HRV nucleic acids in various respiratory tissues [143, 145Bardin, 1994 #110, 146]; nucleic acid sequence-based amplification (NASBA), which directly amplifies HRV RNA [147, 148]; DNA microarray called virochip, that detects all known respiratory viruses [149]; and a new technology coupling nucleic acid amplification to high-performance electrospray ionization mass spectrometry and base-composition analysis [150]. These new generic detection tools will likely result in the discovery of previously unknown HRV strains in the future. ANTIVIRAL DRUGS AGAINST HRVS The availability of efficient antiviral drugs against rhinoviruses is certainly of interest given the high frequency of common colds and the importance of more severe HRV infections in selected populations (see clinical manifestations). The first molecule that was studied to fight HRV infections was intra nasal IFN-α, administred intra-nasally, that diminished symptoms caused by experimentally inoculated HRV in human volunteers [151, 152]. However, as IFN itself causes symptoms similar to those of the common cold, the clinical interest of this agent is limited. Recombinant soluble 39

45 ICAM-1 also somewhat reduced the severity of experimental HRV colds [153]. The costs and limited efficacy related to these agents hampered their development. The design of molecules able to inhibit receptor binding resulted in the elaboration of pleconaril, which irreversibly binds the hydrophobic pocket situated in the VP1 canyon [154]. This agent proved to be active in vitro against most HRV types, with variable efficacy however, some HRV types displaying inherent resistance [155]. In vivo, it was able to reduce the severity and duration of symptoms related to HRV colds [154]. Pleconaril also exhibits activity against HEVs, in vitro and in vivo [156], and was promising in the setting of the treatment of severe HEV infections [157]. However, the drug was shown to select resistant quasispecies. Furthermore, pleconaril was associated with induction of cytochrome P-450 3A enzymes [154], and it was rejected by the US Food and Drug Administration (FDA) for safety issues. VP1 remains a potential antiviral target and research focused on the development of drugs inhibiting receptor binding is ongoing. Another antiviral target includes the viral 3C protease. A potent 3C protease inhibitor, rupintrivir, exhibited high activity against both HRVs and HEVs and intranasal administration of this compound reduced symptoms related to HRV infections after inoculation of healthy volunteers [158]. This effect was not reproducible in naturally infected subjects however, which impeded further drug development. An orally administered 3C protease inhibitor, compound 1, was subsequently designed. This molecule displayed broad anti-hrv activity, but its glutamine aldehyde had the tendency to cyclize with a side chain of the compound, resulting in loss of antiviral activity [159]. Similar 3C inhibitors devoid of this flaw are currently being studied. Other strategies being developed include targeting the viral 2C protein, the viral 3A protein, the viral 3D polymerase, as well as host cell factors [160]. 40

46 In summary, no antiviral agent against HRVs is available so far. Factors impeding the development of such drugs include the necessity for the compound to be active against a large variety of HRV types, the emergence of resistance, and the requirement for the quasi absence of adverse effects. Indeed, the potential adverse effects of an anti-hrv agent should be significantly less harmful than the common cold itself for the drug to elicit interest. Another difficulty resides in the fact that an antiviral drug needs to be administered very early in the course of an HRV infection in order to be clinically effective, as the duration of an HRV infection is short. However, the finding that HRVs are related to asthmatic and COPD exacerbations as well as to chronic infections in immunosuppressed patients raises interest in pursuing research in the field of anti-hrv antivirals. ANTI-HRV VACCINES The development of an anti-hrv vaccine would be of high medical interest, given the frequency of HRV infections and the complications resulting from HRV infections in selected populations. However, the development of such a vaccine is limited by the fact that it should protect against all known HRV types. A recent study brought some hope in this area of research. Edlmayr et al [161] were able to create an anti-hrv vaccine by immunizing mice with recombinant HRV-B14 and HRV-A89 VP1 proteins. The resulting antisera were able to neutralize several HRV types in vitro. However, not all HRV types were shown to be neutralized by these experimental vaccines, further illustrating the difficulty related to the elaboration of an effective anti-hrv vaccine. Even if such a vaccine was available, its efficiency would likely be only temporary, given the possibility of an antigenic drift related to the high mutation rate of these viruses. 41

47 Objectives EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION Though less frequently than for HEVs, recombination events are observed in HRVs, and probably contribute to their evolution. Such events occur among members of the same HRV species, except for ancient HRV-C/HRV-A recombination at the 5 UTR-ORF junction. The aim of this work was to analyse to what extent HRVs could recombine among members of the same species or among members of different species, and in which regions of the genome such recombination could generate competent viral genomes. Secondary goals were 1) to map preferential recombination sites and 2) to study the phenotypes of interspecies HRV recombinants. We used two distinct experimental approaches to achieve these goals. The first involved the design of chimeric HRV genomes generated by molecular cloning methods, whose ability to yield infectious viral particles was assessed by reverse genetics. The second approach relied on non replicative recombination. Constructs representing defective but complementary viruses were co-transfected in cells. Recovery of infectious viruses from these transfected RNAs implied that viable recombination had occurred, and recombination sites were mapped by genome sequencing. 42

48 RHINOVIRUS RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS As mentioned in the introduction, HRV RNA quantification in respiratory specimens is of clinical interest. We assessed the feasibility of this procedure by testing HRV-positive nasal washes and bronchoalveolar lavages with a one-step real-time RT-PCR assay designed in our laboratory, and experimentally estimated the implication of different parameters in the quantification error margin. Materials and methods Materials and methods used for experiments corresponding to results, parts 1.2. and 1.3., are described here whereas material and methods corresponding to results, part 1.1. and part 2., are described in the corresponding publications [162, 163]. PLASMIDS AND CONSTRUCTS Primers used for the construction of chimeric HRV genomes of partially deleted HRV genomes are listed in Table 2. 43

49 Table 2. Primers used for chimeric and deleted HRV contructs. 1. Primer name RV 38 RV39 RV40 RV41 RV42 RV44 RV46 RV47 RV 50 RV 52 RV 53 RV 54 RV 55 Primer sense Forward GGTCACAGCTTGTCTGTAAGCGG 5'-3' sequence Reverse TGTATCCAGTTTTAACCTATAGTGAGTCGTATTAATTTCGCGGG Forward TACGACTCACTATAGGTTAAAACTGGATACAGGTTGTTCCCACCTG Reverse CTGAGATATCCATGGTACCACAATG Forward GATTACAATCAACATGTTCCATTGTGG Reverse GACATTACTGCTTCTTTAAGTTCCACATC Forward CAAGGATCCAAGACTTGATGTGG Reverse CGGGGACGCGTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTATGATA GTCAATTCTAAACAACTTGAATACCATATC Forward GTTATCATGGGCGCCCAAGTATCTAGACAG Forward TGTGCTGAAGAACAAGGACCAATTAGTGACTATGTAACACAATTAG Reverse ACTAATTGGTCCTTGTTCTTCAGCACAGTGAAAGTG Forward AGCACACACACAAGGAATCACTGATTACATACACATGC Reverse AATCAGTGATTCCTTGTGTGTGTGCTAGATTGCTATATGG RV 56 Reverse TCCAATCGATAAATTTGGATATTTTG RV 57 Reverse CTGAAAGGCCTCTGGAGACTG RV 77 Reverse AATCAGTGATTCCTTGTTCTTCGGCGCAATGGAAATGTC RV 78 Forward CGCCGAAGAACAAGGAATCACTGATTACATACACATGC RV 83 Forward GTATATCCCTCCTTATGAGTTGTTAAGACATGAATGGTATG RV 84 Reverse CATACCATTCATGTCTTAACAACTCATAAGGAGGGATATAC RV 87 Forward GAAAAATTCTAGATATAGAACTGATAAATAATTAATTCGTTTGTTTTATAAAAAAAAA RV 88 Reverse TTTTTTTTTATAAAACAAACGAATTAATTATTTATCAGTTCTATATCTAGAATTTTTC RV 97 Forward GATCTGGACAAACTTAAGATAATAGCATATGGTGATGATG RV 98 Reverse CATCATCACCATATGCTATTATCTTAAGTTTGTCCAGATC RV 99 Forward ATGGGCGCCCAGGTTTCTACACAG RV 100 Reverse AATCAGTGATTCCCTGTTCCTCTGCGATACACTCC RV 101 Forward CGCAGAGGAACAGGGAATCACTGATTACATACACATGC Sequence tags are shown in italic. Nucleotide changes introduced by site-directed mutagenesis are shown in bold. 44

50 Construction of the chimeric HRV P1-2A genomes Parental genomes Two parental plasmids were used as vectors to create four chimeric HRV P1-2A constructs. The first was the pr16.11 plasmid (kindly provided by W.-M. Lee, University of Wisconsin, USA) containing a full HRV-A16 genome. We introduced an SfoI site by site-directed mutagenesis at the 5 UTR VP4 junction, so that the SfoI site spanned codons 2 and 3 of the HRV-A16 ORF, as described before [162]. The resulting plasmid was designated QCpHRV- A16. The second plasmid was a pwr3.26 plasmid containing a complete HRV-C11 genome (GenBank accession no. EU840952), called pwr3.26-hrv-c11. This construct was generated in the laboratory by PCR amplification and subcloning of the complete genome of an HRV-C clinical strain collected from a child with pericarditis (see below)[164]. Since this genome was the 11 th HRV-C genome whose complete sequence was released in GenBank, it was classified as HRV-C11 [69]. Of note, for unknown reasons, this infectious clone never gave rise to replication competent virus. The global pwr3.26-hrv-c11cloning strategy consisted of the replacement HRV-B14 genome encoded in an infectious clone, pwr3.26-hrv-14 (kindly provided by W.-M. Lee, University of Wisconsin, USA), by the HRV-C11 genome. HRV-C11 complete genome was amplified and cloned from the clinical sample in three overlapping PCR products. The first PCR product was generated by fusion of two overlapping PCRs. PCR1 (forward primer RV38 containing a SalI restriction site and reverse primer RV39) started upstream of the HRV-B14 first nucleotides and ended with RV39 primer, tagged with the first nucleotides of the HRV- C11 genome sequence. The resulting PCR product was 322 bp long. PCR2 covered the first 45

51 2130 nt of the HRV-C11 genome, starting with primer RV40, tagged with pw3.26 vector sequence, and ending with RV41, harboring an EcoRV restriction site. These two PCR products were fused together by PCR, using primers RV38 and RV40. In this process, the two overlapping PCR products were mixed together without primers in a 5 cycle PCR, allowing the DNA polymerase to elongate the fusion PCR product. Primers RV38 and RV40 were then added for 35 subsequent PCR cycles, leading to the amplification of the complete fusion PCR (Product 1). Product 1 was subcloned into a pcr2.1-topo TA vector using the TOPO TA Cloning Kit (Invitrogen) according to the manufacturer s instructions. Resulting plasmids were amplified in OneShot TOP10 Competent cells (Invitrogen) according the manufacturer s instructions. Minipreps were performed with a NucleoSpin Plasmid Miniprep kit (Macherey- Nagel). Correct cloning was assessed by restriction analysis with SalI and EcoRV enzymes. Product 2 was generated by a PCR performed on the HRV-C11 cdna using primers RV42 and RV44. This PCR product was 3777 bp long and was flanked by an EcoRV restriction site at its 5 end and a BamHI restriction site at its 3 end. It was subcloned into a pcr2.1-topo TA vector as described above, and correct cloning was assessed by restriction analysis with EcoRV and BamHI enzymes. Product 3 was generated by a PCR performed on the HRV-C11 cdna using primers RV46 and RV47, harboring the HRV-C11 poly(a) tail sequence and a tag including a 10 bp long pwr3.26 sequence downstream the HRV-B14 genome containing an MluI resitriction site. This PCR product was 1340 bp long and was flanked by a BamHI restriction site at its 5 end and a MluI restriction site at its 3 end. It was subcloned into a pcr2.1-topo TA vector as described above, and correct cloning was assessed by digestion with BamHI and MluI restriction enzymes. The pwr3.26-hrv-14 plasmid and product 1 subcloned into the pcr2.1-topo TA vector were both digested with SalI and EcoRV enzymes. The digested pwr3.26-hrv-14 vector 46

52 and segment 1 were gel purified and ligated during 5 minutes using T4 fast ligase at room temperature (RT). The ligation product, named pwr3.26-step1, was amplified in in OneShot TOP10 competent cells (Invitrogen) according the manufacturer s instructions and correct cloning was assessed by restriction analysis with SalI and EcoRV enzymes. The pwr3.26-step1 plasmid and product 2 subcloned into the pcr2.1-topo TA vector were both digested with EcoRV and BamHI enzymes. The digested pwr3.26-step1 vector and product 2 were gel purified and ligated overnight using T4 ligase (Promega) at 16 C. The ligation product, named pwr3.26-step2, was amplified in OneShot TOP10 competent cells (Invitrogen) according the manufacturer s instructions and correct cloning was assessed by restriction analysis with EcoRV and BamHI enzymes. The pwr3.26-step2 plasmid and product 3 subcloned into the pcr2.1-topo TA vector were both digested with BamHI and MluI enzymes. The digested pwr3.26-step2 vector and segment 3 were gel purified and ligated overnight using T4 ligase (Promega) at 16 C. The ligation product, named pwr3.26-hrv-c11, was amplified in OneShot TOP10 competent cells (Invitrogen) according the manufacturer s instructions, and correct cloning was assessed by restriction analysis with BamHI and MluI enzymes. The pwr3.26-hrv-c11 sequence was verified both by Sanger sequencing and high throughput sequencing (Fasteris). HRV-A16/HRV-A81 P1-2A construct (Figure 14) HRV-A81 (ATCC strain) was propagated in Hela Ohio cells. RNA was extracted from the infected cell supernatant, reverse transcribed (see below), and the P1-2A region (nucleotides 618 to 3602) was amplified by PCR using primers RV50, which contains a SfoI restriction site, and RV77, the latter harboring an HRV-A16 sequence tag (shown in italic in Table 2). 47

53 An adjacent downstream HRV-A16 sequence was PCR-amplified from the QCpHRV-A16 plasmid using primers RV78, which harbors an HRB-A81 sequence tag (shown in italic in Table 2), and RV56, which contains a ClaI restriction site. The two resulting PCR products were fused together using primers RV50 and RV56 according to the fusion PCR method described above. The fusion PCR was subcloned into a pcr2.1-topo TA vector, and correct cloning was assessed by restriction analysis with SfoI and ClaI enzymes and subsequent sequencing. The SfoI and ClaI digested subcloned insert was then cloned into the SfoI and ClaI digested QCpHRV-A16 vector. HRV-A16/HRV-B14 P1-2A construct (Figure 14) The P1-2A region of an HRV-B14 genome derived from the pwr3.26-hrv-14 plasmid (nucleotides 629 to 3634) was amplified by PCR using primers RV99 in which a thymine was replaced by a cytosine residue, creating an SfoI restriction site (bold C in Table 2), and RV100, harboring an HRV-A16 sequence tag (in italic in Table 2). An adjacent downstream HRV-A16 sequence was PCR-amplified from the QCpHRV-A16 plasmid using primers RV101, which harbors an HRV-B14 sequence tag (in italic in Table 2), and RV56, which contains a ClaI restriction site. The two resulting PCR products were fused together using primers RV99 and RV56 according by fusion PCR. The fusion PCR product was subcloned into a pcr2.1-topo TA vector as described above, and correct cloning was assessed by restriction analysis with SfoI and ClaI enzymes and subsequent sequencing. The SfoI and ClaI subcloned purified insert was then cloned into the SfoI and ClaI digested QCpHRV-A16 vector. 48

54 HRV-A16/HRV-C11 P1-2A construct (Figure 14) The P1-2A region of HRV-C11 (nucleotides 619 to 3570) was amplified by PCR from the pwr3.26-hrv-c11 plasmid using primers RV38 and RV55, the latter harboring an HRV- A16 sequence tag (in italic in Table 2). An adjacent downstream HRV-A16 sequence was PCR-amplified from the QCpHRV-A16 plasmid using primers RV54, which harbors an HRV-C11 sequence tag (in italic in Table 2), and RV56, which contains a ClaI restriction site. The two resulting PCR products were fused together using primers RV38 and RV56 by fusion PCR. The fusion PCR was subcloned into a pcr2.1-topo TA vector, and correct cloning was assessed by restriction analysis with SfoI and ClaI enzymes and subsequent sequencing. The SfoI and ClaI subcloned digested insert was then cloned into the SfoI and ClaI digested QCpHRV-A16 vector. HRV-C11/HRV-A16 P1-2A construct (Figure 14) The P1-2A region of HRV-A16 (nucleotides 619 to 3603) was amplified by PCR from the QCpHRV-A16 plasmid using primers RV50 and RV53 (Table 2), the latter harboring an HRV-A16 sequence tag. An adjacent downstream HRV-C11 sequence was PCR-amplified from the pwr3.26-hrv-c11 plasmid using primers RV52, which harbors an HRV-A16 sequence tag (in italic in Table 2), and RV57, which contains a StuI restriction site. The two resulting PCR products were fused together using primers RV50 and RV57 by fusion PCR. The fusion PCR was subcloned into a pcr2.1-topo TA vector as described above, and correct cloning was assessed by restriction analysis with SfoI and StuI enzymes and subsequent sequencing. The SfoI and StuI subcloned digested insert was then cloned into the SfoI and StuI digested pwr3.26-hrv-c11. 49

55 Construction of the deleted HRV parental genomes used in non replicative recombination experiments Partially deleted HRV genomes were created using restriction digestion, fragment extraction and self-ligation from three plasmids harboring an HRV-A16 genome, an HRV-A39 genome and an HRV-C11 genome, respectively. HRV-A16 del5 UTR( ) and HRV-A16 del5 UTR(1-434) constructs (Figure 16) HpaI- and SfoI- digested QCpHRV-A16 plasmid was gel-purified and self ligated, resulting in the HRV-A16 del5 UTR ( ) construct, lacking the 330 last nucleotides of the 5 UTR (Figure). A second version of this construct was designed, named HRV-A16 del5 UTR ( ) 3 marker, harboring a 3 sequence marker that allowed identification of HRV-A16/HRV-A16 recombinants in co-transfection experiments. For this purpose, 4 silent nucleotide changes were made in a 6 nucleotides stretch situated in the 3 end of the 3D region ( ), using site-directed mutagenesis (QuickChange II XL Site-Directed Mutagenesis kit; Stratagene) with primers RV83 and RV84 (the introduced nucleotide changes are shown in bold in Table 2), according to manufacturer s instruction. The HRV-A16 del5 UTR (1-434) construct was elaborated as follows: a DNA sequence comprising the 191 last nucleotides of the HRV-A16 5 UTR and the 15 first nucleotides of the HRV-A16 ORF was created by gene synthesis (Eurofins), flanked by a SalI restriction site at the 5 end and by a SfoI restriction site at the 3 end and cloned into a TOPO vector; the 50

56 latter was digested with SalI and SfoI; the insert was gel-purified and cloned into SalI- and SfoI-digested QCpHRV-A16 plasmid. Again, a second version of this construct was designed, named HRV-A16 del5 UTR (1-434) 3 marker, harboring a sequence marker in the 3 UTR that allowed identification of HRV- A16/HRV-A 16 recombinants in co-transfection experiments. For this purpose, 5 nucleotides from the HRV-A16 3 UTR sequence were replaced by site-directed mutagenesis (QuickChange II XL Site-Directed Mutagenesis kit; Stratagene) with the corresponding sequence of HRV-A81 using primers RV87 and RV88 (Table 2) according to manufacturer s instruction. This manipulation thus allowed to introduce five nucleotide substitutions which were used as a 3 end sequence marker without disrupting the 3 UTR function. HRV-A16 del3 ( ) construct (Figure 16) The phrv-a16 plasmid was double digested with SmaI and self ligated, thereby removing the 3 end of the HRV-A16 genome (from nucleotide 5284 to the end of the poly(a) tail), including the entire 3D gene, coding for the viral RNA polymerase. HRV-A39 del3 ( ) construct (Figure 16) An additional AflII restriction site was introduced into the 3D region of the phrv-39 plasmid at position 6664 by site-directed mutagenesis (QuikChange II XL Site-Directed Mutagenesis kit; Stratagene) with primers RV 97 and RV 98 (Table 2). The modified phrv-a39 plasmid, named QCpHRV-A39, was double digested with AflII and self-ligated, yielding the HRV-A39 del3 ( ) genome. 51

57 HRV-C11 del3 ( ) construct (Figure 16) After double digestion with MluI and AflII, the pwr3.26-hrv-c11 vector was gel-purified, treated with T4 DNA Polymerase for blunting, and self-ligated. The resulting HRV-C11 del3 ( ) genome lacks the second half of the 3D region as well as the whole 3 UTR. RNA EXTRACTION AND REVERSE TRANSCRIPTION RNA was extracted with a NucliSens easymag magnetic beads system (biomérieux) according to the manufacturer s instructions, reverse transcribed with Superscript II (Invitrogen) and random hexamer primers (Roche) as previously described [162]. IN VITRO TRANSCRIPTION AND TRANSFECTION In vitro transcription was performed as described in [162]. For single transfections, 2 µg of in vitro-transcribed RNA per well was transfected in six-well plates containing % confluent HeLa Ohio cells grown on coverslips using a TransMessenger Transfection Reagent kit (Qiagen), as described previously [18]. The deleted HRV genomes were co-transfected using 1 µg of each in vitro-transcribed RNA per well. CELL CULTURE HeLa Ohio cells (kindly provided by F. H. Hayden, University of Virginia, USA) were grown in Eagle s minimum essential medium (Lonza) supplemented with 2 mm L-glutamine, 1 µg 52

58 amphotericin ml 1, 100 µg gentamicin ml 1, 20 µg vancomycin ml 1 and 10 % FCS at 37 C in a 5 % CO 2 atmosphere. VIRAL CULTURE The HRV-A81 ATCC strain and chimeric HRVs resulting from transfection and cotransfection experiments were propagated in HeLa Ohio cells in McCoy's 5A Medium 2% FCS incubated at 33 C. IMMUNOFLUORESCENCE At different times post transfection, cells were washed twice with PBS lacking Ca 2+ and Mg 2+ (PBS ) and fixed for 20 min in methanol : acetone (1: 1) at 20 C. Cells were air dried for a few minutes at room temperature before incubation with the primary mouse antibody, J2 monoclonal antibody (mab J2) (Engscicons), directed against double-stranded RNA (dsrna) longer than 40 bp [165], diluted 1 : 500 in PBS /1% BSA, for 45 min at 37 C. After intensive washing with PBS, FITC-conjugated anti-mouse IgG antibody (Light Diagnostics) was added to the cells for 45 min at 37 C in the dark. Cells were then washed three times with PBS and stained with DAPI for 5 min at room temperature. After a final rinse with PBS, the coverslips were mounted in Fluoroprep mounting medium (biomérieux). Mounted samples were then analysed by fluorescence microscopy. 53

59 SEQUENCING AND MAPPING OF RECOMBINATION SITES Cells transfected with chimeric genomes or co-transfected with deleted genomes were observed at different time post transfection. In the presence of specific cytopathic effect, RNA was extracted from the infected cell supernatant and reverse transcribed as described above. Selected parts of recombinant HRV genomes were amplified by PCR and sequenced using the Sanger method. Chromatograms obtained using an ABI Prism 3130XL DNA Sequencer (Applied Biosystems) were analyzed with the Geneious Pro software (Biomatters Ltd). Regarding the chimeric HRV-A16/HRV-A39 P12A virus, a PCR product encompassing the chimeric 2A-2B junction was sequenced to confirm the chimeric nature of the growth competent virus. Regarding non replicative HRV-A16/HRV-A39 RNA recombination, PCR products amplifying adjacent genomic regions, starting from the 3 end of the genome were produced using degenerate primers allowing the amplification of different HRV types, until the recombination site was identified by sequencing. 54

60 Results EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION Chimeric HRV genomes generated by molecular cloning methods Artificially engineered HRV interspecies recombinants at the 5 UTR-ORF junction are viable and replication competent: Article 1 Experimental human rhinovirus and enterovirus interspecies recombination. Schibler M, Gerlach D, Martinez Y, Belle SV, Turin L, Kaiser L, Tapparel C. J Gen Virol Jan;93(Pt 1):

61 Journal of General Virology (2012), 93, DOI /vir Experimental human rhinovirus and enterovirus interspecies recombination Manuel Schibler, 1 Daniel Gerlach, 2 Yannick Martinez, 3 Sandra Van Belle, 1 Lara Turin, 1 Laurent Kaiser 1 and Caroline Tapparel 1 Correspondence Manuel Schibler manuel.schibler@hcuge.ch 1 Laboratory of Virology, Division of Infectious Diseases and Division of Laboratory Medicine, University of Geneva Hospitals, 4 Rue Gabrielle-Perret-Gentil, 1211 Geneva 14, Switzerland 2 Research Institute of Molecular Pathology (IMP), Dr Bohr-Gasse 7, A-1030 Vienna, Austria 3 Department of Pathology and Immunology, Faculty of Medicine, University of Geneva, Geneva, Switzerland Received 1 July 2011 Accepted 16 September 2011 Human rhinoviruses (HRVs) and enteroviruses (HEVs), two important human pathogens, are nonenveloped, positive-sense RNA viruses of the genus Enterovirus within the family Picornaviridae. Intraspecies recombination is known as a driving force for enterovirus and, to a lesser extent, rhinovirus evolution. Interspecies recombination is much less frequent among circulating strains, and supporting evidence for such recombination is limited to ancestral events, as shown by recent phylogenetic analyses reporting ancient HRV-A/HRV-C, HEV-A/HEV-C and HEV-A/HEV-D recombination mainly at the 59-untranslated region (59 UTR) polyprotein junction. In this study, chimeric genomes were artificially generated using the 59 UTR from two different clinical HRV-C strains (HRV-Ca and HRV-Cc), an HRV-B strain (HRV-B37) and an HEV-A strain (HEV-A71), and the remaining part of the genome from an HRV-A strain (HRV-A16). Whilst the chimeric viruses were easily propagated in cell culture, the wild-type HRV-A16 retained a replication advantage, both individually and in competition experiments. Assessment of protein synthesis ability did not show a correlation between translation and replication efficiencies. These results reflect the interchangeability of the 59 UTR, including its functional RNA structural elements implicated in both genome translation and replication among different enterovirus species. The 59 UTR polyprotein junction therefore represents a theoretic interspecies recombination breakpoint. This recombination potential is probably restricted by the need for co-infection opportunities and the requirement for the progeny chimera to outcompete the parental genomes fitness, explaining the rare occurrence of such events in vivo. INTRODUCTION Human rhinoviruses (HRVs) and human enteroviruses (HEVs) are non-enveloped, positive-sense RNA viruses and members of the genus Enterovirus, the largest genus in the family Picornaviridae. HRVs currently consist of 151 proposed types (Knowles, 2011; Simmonds et al., 2010), classified into three species (HRV-A, -B and -C), and represent the most common cause of respiratory tract infection in humans (Garbino et al., 2009; Mäkelä et al., 1998; Ruohola et al., 2009). HEVs encompass 107 types classified into four species (HEV-A, -B, -C and -D) and cause various mild to severe clinical manifestations, especially in children, such as hand, foot and mouth disease, meningoencephalitis, poliomyelitis and myopericarditis (Sawyer, 2001). A supplementary table and figure are available with the online version of this paper. HRVs and HEVs share a similar genomic organization, which consists of a 59-untranslated region (59 UTR), a single ORF, a 39 UTR and a poly(a) tail. The ORF encodes a polyprotein that is co-translationally cleaved into four structural viral particle proteins (VP4, VP2, VP3 and VP1) and seven non-structural proteins (2A, 2B, 2C, 3A, 3B, 3C and 3D). The genomic RNA of these viruses also harbours a stem loop structure called the cis-acting replication element (cre), which is essential for replication. The precise location of this cre stem loop structure is constrained to the 2C region for all HEV species, whereas its position varies in the three HRV species and overlaps the 2A (HRV- A), VP1 (HRV-B) and VP2 (HRV-C) regions (Cordey et al., 2008). The enterovirus 59 UTR forms six distinct secondarystructure domains that can be divided into two functional units: a 59 cloverleaf structure (CL) essential for replication (domain I) and an internal ribosome entry site (IRES) G 2012 SGM Printed in Great Britain 93

62 M. Schibler and others (domains II VI) necessary for the cap-independent translation of the polyprotein (Rohll et al., 1994). The CL is further subdivided into four structural domains: stem A, stem loop B, stem loop C and stem loop D. Stem loop B is a known binding site for cellular proteins called poly(rc)-binding proteins (PCBPs), and stem loop D interacts with the viral 3C and 3CD proteases. The CL, cellular PCBPs and viral 3CD are all implicated in the switch from translation to replication (Du et al., 2004; Perera et al., 2007). Although replication regulation has been attributed mainly to the CL, sequences located in the 39 region of the poliovirus (PV) IRES may also be implicated in this process (Borman et al., 1994). Viral IRESs are classified into five types according to primary sequence, secondary structure, location of the initiation codon and activity in different cell types (Racaniello, 2007). HRVs and HEVs both share a loosely structured type I IRES, which is characterized by five key stems (domains II VI) involved in protein binding and interaction with the 43S ribosomal subunit (Palmenberg et al., 2010). HRVs and HEVs are characterized by an important diversity of types, and previously unknown strains are constantly being reported (Arden et al., 2006; McErlean et al., 2007; Smura et al., 2007; Tapparel et al., 2009a; Yozwiak et al., 2010). The main explanations for such an important variability are the high error rate of the viral RNAdependent RNA polymerase and recombination (Domingo & Holland, 1997). Intraspecies recombination events have been extensively described for HEVs and represent an evolutive force for this virus group (Lukashev, 2005; Santti et al., 1999). The enterovirus recombination breakpoints are most frequently reported around the 59 (VP4) and 39 (2AB) ends of the P1 region, whilst they are almost absent in the capsid VP2 VP3 VP1 region (Lukashev et al., 2005; Simmonds & Welch, 2006). Evidence for recombination is more limited among rhinoviruses. Few recombination events have been reported for circulating HRV strains (Tapparel et al., 2009b), with breakpoints located at the 39 end of the 59 UTR and the 59 end of the 3C gene. However, analysis of the full-length sequences of all known HRV serotypes suggests that some serotypes have actually resulted from ancient intraspecies recombination (Palmenberg et al., 2009). Natural interspecies recombination among HEVs may have occurred in the past. Strains resulting from putative 59 UTR exchange between HEV-C and HEV-A as well as between HEV-D and HEV-A have been reported recently (Smura et al., 2007). Similarly, Yozwiak and co-workers described a novel HEV-C (EV-109) whose 59 UTR was presumably acquired through recombination with an HEV-A member (Yozwiak et al., 2010). Likewise, recombination between the 59 UTR of HRV-A and the polyprotein of HRV-C was proposed as the mechanism responsible for the generation of the HRV-Ca subgroup exhibiting HRV-A-like 59 UTR sequences (Huang et al., 2009; McIntyre et al., 2010). The remaining HRV-C strains, called HRV-Cc, have 59 UTR sequences that segregate from those from HRV-A, HRV-B and HRV-Ca members. Putative interspecies recombination breakpoints in the 59 UTR have been mapped for HRV-Ca strains: one around position 481, the second around position 565 in the polypyrimidine tract and the third around position 523 within stem loop 5 of the IRES (McIntyre et al., 2010). In the majority of sequences analysed, recombination presumably occurred in either one of the last two recombination hotspots, which are located in highly conserved sequence stretches. These two particular locations may therefore represent preferred sites for other interspecies 59 UTR recombination within members of the genus Enterovirus. Furthermore, some HRV-C strains harbour short HRV-A sequences in their 2A region (Huang et al., 2009; McIntyre et al., 2010). Finally, based on full-genome phylogenetic analysis, we have proposed that ancient recombination events between HRV-A and HEV members gave rise to the HRV-B species (Tapparel et al., 2007). However, based on sequence homology, all the above proposed natural interspecies HEV and HRV recombination events probably occurred between ancestors of the current HEV and HRV circulating strains. Constructed viable interspecies HRV and HEV recombinants have also been reported in the literature. Examples include a chimeric PV1/HRV-14 construct in which the 59 UTR was derived from PV-1 and the remainder of the genome from HRV-14 (Todd et al., 1997), a chimeric HRV-2/PV-1 in which the PV-1 IRES is replaced by an HRV-2 IRES (Gromeier et al., 1996), and a chimeric coxsackievirus 4 (CV-B4)/PV-3 consisting of a PV-3 genome in which the 59 UTR is derived from CV-B4 (Rohll et al., 1994). This study aimed to explore further the recombination potential of different enterovirus species at the 59 UTR polyprotein junction, and therefore the compatibility between these species at that level. For this purpose, we assessed the viability of chimeras with 59 UTRs originating from different species (HRV-Ca, HRV-Cc, HRV-B and HEV-A) fused to a common HRV-A (HRV-A16) backbone. Each of these recombinant genomes gave rise to infectious viral particles. We also tested and compared the replication, translation and competition abilities of these 59 UTR chimeric viruses. RESULTS Chimera constructs The four chimeric genomes represented in Fig. 1 were constructed using a plasmid encoding the full-length HRV- A16 genome in which the entire 59 UTR was replaced by the 59 UTR originating from an HRV-Ca, HRV-Cc, HRV- B37 or HEV-A71 strain (see Methods). The resulting plasmids were then used to generate virions by transfection of in vitro-transcribed RNA into HeLa Ohio cells. A comparison of the 59 UTR sequence homologies of the viral genomes used in this study is shown in Table 1. The 94 Journal of General Virology 93

63 Experimental HRV and HEV recombination Fig. 1. Schematic representation of the four chimeric genomes and the parental HRV-A16. (a) Each construct s name, 59 UTR origin, common polyprotein and 39 UTR are shown. The target region for the HRV-A16 3D-specific real-time PCR assay is indicated (arrow). HRV-Ca and HEV-A71 59 UTRs displayed the highest (77 %) and lowest (64 %) sequence similarities, respectively, compared with the HRV-A16 59 UTR. As expected, HRV-A and HRV-Ca did not segregate on a 59 UTR-based sequence phylogenetic tree (see Supplementary Fig. S1a, available in JGV Online), reflecting the previously mentioned 59 UTR transfer from the HRV-A species to the HRV-Ca subspecies. Whilst dendrogram trees based on RNA structure mainly reflected the grouping of known species, some species were completely misplaced with regard to the sequence-based trees (Supplementary Fig. S1b). Although we cannot exclude the limitations of current single-sequence ab initio structure prediction programs, they may indicate the high similarity of structure elements among species, which is not visible from a sequence-based comparison. Of note, as for HRV-A and HRV-Ca, HEV species did not segregate correctly based on their 59 UTR sequence. The chimeric derivatives replicate less efficiently than the non-chimeric parental HRV-A16 The replication ability of each chimera was compared with that of HRV-A16 at 2 h, 48 h and three passages after transfection of standardized amounts of in vitro-transcribed RNA. This experiment was performed in duplicate. This comparison was made using a quantitative real-time PCR assay specific for the HRV-A16 3D region, as shown in Fig. 1. The 1-acylglycerol-3-phosphate-O-acyltransferase (AGPAT) housekeeping gene was used for normalization. Of note, this replication ability assessment was based on the assumption that the target RNA measured reflected viral genome amplification and thus replication. Although the amount of RNA measured at 2 h posttransfection was similar for the different constructs, notable differences in replication efficiency were observed after three passages, as shown in Fig. 2(a). At this time point, the non-chimeric HRV-A16 had the highest viral RNA load, followed by the HRV-Cc/A16 and HEV-A71/ A16 chimeras. The lowest viral RNA load was observed for the HRV-Ca/A16 and HRV-B37/A16 chimeras. The same trend was observed by immunofluorescence with an anti- HRV-A16 antibody performed on HeLa Ohio cells at 48 h post-transfection in duplicate. Bioimaging quantification showed that 28, 24, 23, 18 and 13 % of cells transfected with HRV-A16, HEV-A71/A16, HRV-Cc/A16, HRV-Ca/ A16 and HRV-B37/A16 RNA, respectively, were immunofluorescence positive (results not shown). HRV-A16 outcompetes the chimeras in co-transfection assays, except when present at a lower concentration To assess whether the replication advantage observed for the parental HRV-A16 virus was biologically significant, we designed a competition experiment in which normalized Table 1. Percentage similarities between the different 59 UTRs used in this study. Construct 5 UTR HRV-A16 5 UTR HRV-Ca 5 UTR HRV-B37 5 UTR HRV-Cc 59 UTR HE-A UTR HRV-A UTR HRV-Ca UTR HRV-B

64 M. Schibler and others Fig. 2. Replication efficiencies of the chimeric and non-chimeric HRV-A16 genomes. (a) For each derivative, the level of positive-strand RNA was quantified by real-time RT-PCR and normalized to a housekeeping gene at 2 h post-transfection (not shown), 48 h post-transfection and after three amplification passages (3¾48 h) in HeLa Ohio cells. The viral positive-strand RNA level is expressed relative to HRV-A16 RNA measured at 2 h post-transfection (input RNA) and set as a reference (see Methods for calculation). The relative amount of input RNA of the four chimera constructs was comparable to HRV-A16 RNA, set to 1 (mean 1.20, SD 0.16; data not shown). Error bars were calculated from two separate experiments. (b, c) 59 UTR VP2 PCR products were amplified from viruses obtained at 48 h post-transfection (p1) and after three passages (p3) or four passages (p4) after co-transfection into HeLa Ohio cells with equivalent amounts of the four chimeric RNAs and the nonchimeric HRV-A16 RNA (b) or equivalent amounts of each chimeric RNA and a tenfold lower amount of HRV-A16 RNA (c). The identities of the 59 UTRs determined by sequencing are indicated (arrows). These experiments were carried out in triplicate. L, 1 kb ladder. amounts of the four chimeric and the non-chimeric HRV- A16 RNAs were co-transfected in triplicate. A strong cytopathic effect was observed in each sample at 3 days post-transfection. Cell supernatants were passaged three times. Viral cdnas obtained from the first- and thirdpassage supernatants were PCR amplified over approximately 1 kb of the 59 UTR VP2 region with primers exactly matching each of the five derivatives to avoid any PCR bias. PCR performed on the supernatant collected 48 h after transfection (defined as the first passage; Fig. 2b, lane p1) and after the third passage (Fig. 2b, lane p3) yielded two fragments of different sizes and intensities. These PCR products were subcloned and sequenced. For the first passage, eight of ten clones harboured the HRV-A16 59 UTR sequence, whilst the remaining two contained the HEV-A71 59 UTR sequence. This ratio was 9 : 1 after two additional passages. Notably, sequencing of the gel-purified upper band four passages after co-transfection (Fig. 2c) confirmed the HEV-A71 59 UTR identity; the size of HEV-A71 59 UTR is 743 nt versus nt for the HRVs used in this study. These results indicated a tendency of the non-chimeric HRV-A16 virus to outcompete the four chimeric viruses. The experiment was repeated with a tenfold dilution of HRV-A16 RNA, as shown in Fig. 2(c). Again, the PCR product was subcloned and sequenced. For the first passage, five clones harboured the HEV-A71 59 UTR sequence, three harboured the HRV-B37 59 UTR sequence 96 Journal of General Virology 93

65 Experimental HRV and HEV recombination and one harboured the HRV-Cc 59 UTR sequence. This mixed population suggested that, after one passage, HEV- A71 presented a slight replication advantage over the other chimeras and the diluted HRV-A16. This advantage became important after four passages, as eight clones contained the HEV-A71 59 UTR sequence and two contained the HRV-Cc 59 UTR sequence. These results showed that, when HRV-A16 was present at a lower concentration, it was unable to outcompete the recombinant viruses. In addition, these data tended to confirm that, aside from the non-chimeric HRV-A16, HEV-A71/A16 and HRV-Cc/A16 replicated the most efficiently. There is no correlation between replication and translation efficiencies The translation abilities of the chimeras and parental HRV- A16 were assessed using a cell infection assay. For each virus, a 50 % cell-culture infective dose (CCID 50 )of10 5 ml 21 was used to inoculate HeLa Ohio cells. To minimize the impact of replication on translation efficiency, a Western blot was performed on cell lysates at early time points (2, 6 and 9 h) after infection, and the amount of protein was compared with the amount of normalized intracellular positive-strand RNA (the template for translation) measured by quantitative real-time PCR (Fig. 3 and data not shown). RNA levels measured at 2 and 6 h postinfection (p.i.) were comparable among the five viruses, and increased amounts were observed only after 9 h, suggesting that replication started between 6 and 9 h for all derivatives. Furthermore, amplification of the positivestrand RNA matrix appeared to be necessary for protein detection as no viral proteins were visible by Western blotting before 9 h (data not shown). Both Western blot and RNA quantification results were reproducible with biological replicates. Overall, viral protein and positive-strand RNA amounts measured at 9 h p.i. correlated well except for the HRV-Cc/ A16 and HRV-Ca/A16 chimeras. HRV-Cc/A16 displayed low levels of protein synthesis, despite intermediate positivestrand RNA levels at 9 h p.i. (Fig. 3) and high replication levels afterwards (Fig. 2a). However, HRV-Ca/A16 chimera translation ability was equivalent to that of HEV-A71/A16 and HRV-A16 (Fig. 3a), whilst its replication potential was notably lower at later time points (Fig. 2a). Fig. 3. Comparative translation efficiencies of the chimeric genomes and the non-chimeric HRV-A16. (a) Western blot analysis at 9 h p.i. of HeLa Ohio cell lysates with an anti-hrv-16 VP2 mab (which also detects the VP0 precursor) and an anti-actin antibody. Protein identities are indicated on the right. (b) For each derivative, the level of positive-sense RNA was quantified by real-time RT-PCR and normalized to a housekeeping gene at 2, 6 and 9 h p.i. The viral positive-strand RNA level is expressed relative to HRV-A16 RNA measured at 2 h p.i. (input RNA) and set as a reference (see Methods for calculation). RNA samples monitored at 9 h p.i. were extracted from the same cell lysates as those used for Western blotting (a). Results are shown as means±sd derived from two separate experiments. 97

66 M. Schibler and others The growth phenotype of the HEV-A71/A16 chimera is defined by its polyprotein rather than its 5 UTR sequence HRVs and HEVs exhibit a different cell tropism and optimal growth temperature in vivo. To assess whether the 59 UTR region might play a role in these different phenotypes, we compared the growth ability of the HEV- A71/A16 chimera with the non-chimeric HEV-A71 and HRV-A16 in different cell types and at different temperatures. When the chimeric RNA was transfected into HeLa Ohio cells, propagation of the resulting virus was similar to that of HRV-A16 and much stronger than that of HEV- A71 (data not shown). In addition, as for HRV-A16, the optimal growth temperature was 33 uc. When transfected into Vero cells, a cell line typically used for HEV culture, there was no visible cytopathic effect for the HEV-A71/16 chimera or for HRV-A16 at 33 or 37 uc, whereas the HEV- A71 virus replicated efficiently (data not shown). Overall, the viral culture phenotype of the chimera containing the HEV-A71 59 UTR and HRV-A16 polyprotein was similar to that of the parental rhinovirus. DISCUSSION Recombination is a well-described driving force for picornavirus evolution. This phenomenon is reported more frequently for HEVs than for HRVs (Lukashev, 2005; McIntyre et al., 2010; Santti et al., 1999) and occurs mainly within members of the same species. Interspecies recombination events have rarely been reported and, based on sequence similarity, these recombinations have been described mainly as ancient events, possibly at the origin of new HEV types, subspecies or species (Huang et al., 2009; McIntyre et al., 2010; Smura et al., 2007; Tapparel et al., 2007; Yozwiak et al., 2010). The aim of our study was to test the interspecies recombination potential at the 59 UTR ORF junction among members of the genus Enterovirus. Our results revealed that the 59 UTRs of selected representatives of HRV-Cc, HRV-Ca, HRV-B and HEV-A were all functional in the context of an HRV-A (HRV-A16) genome. This implies that the 59 UTR including the IRES and CL cis-acting elements of each of these species are recognized efficiently by HRV-A16 proteins, resulting in productive translation and replication. This functional compatibility hints at an interspecies recombination potential between the 59 UTR and the rest of the genome within members of the genus Enterovirus. This interspecies recombination opportunity may, however, be limited by several factors. First, two different HRVs and/or HEVs have to co-infect a target cell in the same time frame. This may seem particularly unlikely with regard to HEV/HRV recombination, as HEVs are known for their enteric tropism. However, several HEV strains exhibit a respiratory tropism, and the simultaneous presence of an HEV and an HRV genome in the same clinical specimens has been documented (Lu et al., 2008). Regarding HRV/HRV co-infection, tropism is not a limiting factor, and we have observed such events in the past following the use of separate HRV-A and HRV-B real-time PCR assays to screen respiratory samples. Although these observations do not imply that the viruses co-detected in the same anatomical site are also present in the same cell, reports of intraspecies HRV recombinants suggest that such singlecell co-infection indeed occurs. Secondly, to outcompete the non-chimeric parental strains present in superior amounts in the co-infected host and to spread in the population, recombinant viruses need to be fitter than the non-chimeric viruses. In the setting of our competition experiments, the non-chimeric HRV-A16 virus was revealed to be fitter than the recombinants. This may be explained by suboptimal interactions between the 59 UTR and other parts of the genome, as well as viral proteins. These interactions have probably been optimized over time by co-evolution thanks to both purifying and diversifying selective pressure (Kistler et al., 2007), and well-adapted strains may be difficult to outcompete. However, our competition assays were performed in HeLa Ohio cells, and it is not certain whether these considerations are valid for natural recombinants occurring in vivo. Productive interspecies recombination within the polyprotein region may be even more complicated, as chimeric polyproteins may harbour protein cleavage sites incompatible with 2A and 3C proteases, and cre motifs, situated at different positions among different species, may be deleted or duplicated. Surprisingly, we observed that the recombinant virus with the most divergent 59 UTR sequence and the second most divergent 59 UTR structure with regard to HRV-A16 (Supplementary Fig. S1), namely HEV-A71/A16 (64 % sequence similarity to HRV-A16), was the second fittest virus, both in independent replication assays and in competition experiments, and displayed an efficient translation potential. In contrast, HRV-Ca/A16 and HRV-B37/ A16 replicated with low efficiency. Interestingly, these two chimeras clustered in the structure-based RNA tree shown in Supplementary Fig. S1(b), suggesting that their secondary RNA structure may drive suboptimal replication and virus propagation in the HRV-A16 background. However, as mentioned earlier, structural predictions made with current programs should be interpreted carefully. We also demonstrated that there was no strict correlation between the translation and replication efficiencies of our chimeras. Indeed, HRV-Cc/A16 displayed a replication pattern almost equivalent to that of the parental HRV-A16, despite a notably lower translation ability, whereas HRV-Ca/A16 replicated markedly less, despite exhibiting a translation ability similar to that of HRV-A16. Finally, we tested the viral culture phenotype of the HEV- A71/A16 chimera. Its optimal replication temperature and cellular tropism closely resembled those of the wild-type HRV-A16 and differed dramatically from those of HEV- A71, indicating that these parameters are not defined 98 Journal of General Virology 93

67 Experimental HRV and HEV recombination by the 59 UTR part of the viral genome. Although the pathogenicity of such interspecies recombinants in vivo is hard to predict, our results suggest that most of the virus phenotype relies on the polyprotein rather than on the 59 UTR sequence. In summary, we have shown experimentally that the 59 UTRs are functionally interchangeable between selected enterovirus species, and that propagation of such chimeras is limited by competition between parent and progeny in cell culture. Our data support phylogenetic analyses indicating ancient interspecies recombination among HRVs and HEVs, as well as between them. Similar recombination events may occur in the future, further contributing to the variability of these viruses. METHODS Construction of the chimeric HRV-A16 derivatives. The pr16.11 plasmid (kindly provided by W.-M. Lee, University of Wisconsin, USA) containing a full HRV-A16 genome was used to create four chimeras with different 59 UTRs. An SfoI site was introduced by sitedirected mutagenesis (QuikChange II XL Site-Directed Mutagenesis kit; Stratagene) at the 59 UTR VP4 junction with primers RV50 and RV51 (Supplementary Table S1, available in JGV Online ), so that the SfoI site spanned codons 2 and 3 of the HRV-A16 ORF. The resulting plasmid, designated QCpHRV-A16, was then depleted of its 59 UTR by SalI/SfoI digestion to receive the 59 UTR prepared from HRV-Ca, HRV-Cc and HRV-B and HEV-A71, as described below. For the HRV-Ca/A16 construct, a pwr3.26 plasmid containing a complete HRV-C11 genome (GenBank accession no. EU840952) generated in our laboratory was digested with SalI and SfoI. The HRV-Ca 59 UTR was gel purified and cloned into the purified SfoI/ SalI-digested QCpHRV-A16 vector. For all other constructs, PCR was used to amplify 59 UTR products (using Platinum Taq DNA Polymerase High Fidelity; Invitrogen) directly from a clinical specimen (HRV-Cc, GenBank accession no. JN087518), an ATCC strain (HRV-B37, GenBank accession no. EF173423) or an infectious clone (plasmid pev71; BrCr-TR, GenBank accession no. AB204852; kindly provided by A. Minetaro, National Institute of Infectious Diseases, Japan), using a SalI T7 promoter-tagged forward primer and an SfoI-tagged reverse primer (primers RV89 RV96; Supplementary Table S1). The PCR products were subcloned into a pcr2.1-topo vector (Invitrogen) before being digested by SalI and SfoI and cloned into the purified SalI/SfoIdigested QCpHRV-A16 vector. The in vitro-transcribed parental QCpHRV-A16 and pev71 (BrCr- TR) RNA (HEV-A71 in the text) were used as controls in various assays. Cell culture and infection. HeLa Ohio cells (kindly provided by F. H. Hayden, University of Virginia, USA) were grown in Eagle s minimum essential medium (Lonza) supplemented with 2 mm L-glutamine, 1 mg amphotericinml 21,100mg gentamicinml 21,20mg vancomycin ml 21 and 10 % FCS at 37 uc ina5%co 2 atmosphere. Vero cells were grown in Dulbecco s modified Eagle s medium (DMEM; Gibco) supplemented with 2 mm L-glutamine, 100 mg penicillin/streptomycin ml 21, 10 % FCS and 0.2 % NaHCO 3 at 37 uc in a 5 % CO 2 atmosphere. The infection media used were McCoy s 5A Medium (Invitrogen) supplemented with 30 mm MgCl 2,1 mg amphotericin ml 21,20mg vancomycin ml 21,50mg gentamicin ml 21 and 2 % FCS for HeLa Ohio cells, and DMEM (Gibco) supplemented with 2 mm L-glutamine, 100 mg penicillin/streptomycin ml 21,1mg amphotericin ml 21, 100 mg gentamicin ml 21, 10 % FCS, 0.2 % NaHCO 3 and 2 % HEPES for Vero cells. Viral stocks were prepared as follows: RNA derived from HRV-Ca/ A16, HRV-Cc/A16, HRV-B37/A16, HEV-A71/A16 and HRV-A16 constructs was transfected into HeLa Ohio cells (see below). Cells and supernatants were collected at 72 h post-transfection and submitted to three freeze (280 uc)/thaw cycles. The resulting mixture was clarified and distributed into aliquots. The CCID 50 ml 21 was determined for each virus stock using the method of Karber (1931). Virus stocks were diluted in infection medium to 10 5 CCID 50 ml 21 to infect HeLa Ohio cells, as described elsewhere (Cordey et al., 2010). For virus passaging, 10 or 20 ml infected cell supernatant was added directly to the infection medium overlying fresh HeLa Ohio cells seeded in 12- or six-well plates, respectively. Virus was passaged every 48 h. In vitro transcription and transfection. For each construct, 5 10 mg plasmid was linearized at a unique SacI restriction site downstream of the viral 39 poly(a) tail. RNA transcripts were synthesized from their linear templates with a MEGAscript T7 kit (Ambion) for 3 h at 37 uc and purified with an RNeasy Mini kit (Qiagen). In vitro-transcribed RNA was quantified and checked by 0.1 % SDS/1 % agarose gel analysis. For single transfections in six- or 12-well plates, 2 or 1 mg in vitro-transcribed RNA was transfected per well in HeLa Ohio or Vero cells using a TransMessenger Transfection Reagent kit (Qiagen), as described previously (Cordey et al., 2008). For co-transfection experiments, 0.4 mg RNA of each construct was used or 0.04 mg for the tenfold-diluted HRV-A16 RNA. RNA extraction, reverse transcription, PCR, subcloning and sequencing. Transfected or infected HeLa Ohio cells were checked daily for cytopathic effect. At selected time points, 200 ml supernatant was used for RNA extraction with a NucliSens easymag magnetic beads system (biomérieux) according to the manufacturer s instructions. Reverse transcription was performed with Superscript II (Invitrogen) and random hexamer primers (Roche). Primers 7 and P1.204 were used to identify the viruses resulting from the competition experiments (Supplementary Table S1). PCR products were purified with Microcon columns (Millipore) before sequencing. Gel purification was performed using a QIAEX II Gel Extraction kit (Qiagen), according to the manufacturer s instructions. PCR products were subcloned into the pcr2.1-topo vector using a TOPO TA Cloning kit (Invitrogen), according to the manufacturer s instructions. Minipreps were performed with a NucleoSpin Plasmid Miniprep kit (Macherey-Nagel) and sequenced with M13 forward and reverse primers. Chromatograms produced with an ABI Prism 3130XL DNA Sequencer (Applied Biosystems) were imported directly for proofreading using Geneious Pro software (Biomatters Ltd). Quantitative real-time RT-PCR. Reverse transcription was performed on RNA extracted from total cell lysates using oligo(dt) primers (Roche) to specifically quantify the plus-strand RNA. cdna was analysed by a real-time PCR assay specific for the HRV-A16 3D region (primers P3.142, P3.144 and probe Pr HRV16 3D; Supplementary Table S1), as well as by an AGPAT housekeeping gene assay (primers HsAGPAT F1, HsAGPAT R1 and probe HsAGPAT pro; Supplementary Table S1) with a TaqMan Universal PCR Master Mix (Applied Biosystems) under the following cycling conditions: 50 uc for 2 min, 95 uc for 10 min, followed by 45 cycles of 95 uc for 15 s and 60 uc for 1 min in a 7500 Applied Biosystems 99

68 M. Schibler and others thermocycler. Results were analysed using the SDS version 1.4 program (Applied Biosystems). Viral amplicon C t values were normalized to those of the housekeeping gene. Relative quantification was calculated using the 2 2DDC t method (Livak & Schmittgen, 2001). The quantitative HRV-A16 3D assay was run using a tenfold dilution series of the pr16.11 plasmid, which was used as a quantitative reference curve for each run. SDS-PAGE analyses and Western blotting. Total cellular extracts, lysed in NP-40 lysis buffer [10 mm NaCl, 50 mm Tris/HCl (ph 8), 0.6 % NP-40] supplemented with the protease inhibitors 2 mm 4-(2- aminoethyl)-benzenesulfonyl (Sigma-Aldrich) and aprotinin (diluted 1 : 1000; Sigma-Aldrich), were analysed by SDS-PAGE (15 % acrylamide). After electrophoresis, the proteins were transferred using a semi-dry system onto PVDF membranes (Millipore). Blots were then incubated with a mouse anti-hrv-a16 mab (kindly provided by W.-M. Lee) and a mouse anti-actin mab (Millipore), followed by a goat anti-mouse HRP-conjugated secondary antibody (Bio-Rad). Protein detection was performed using an enhanced chemiluminescence system (Amersham Biosciences). Immunofluorescence. At 48 h post-transfection, cells were washed twice with PBS lacking Ca 2+ and Mg 2+ (PBS 22 ) and fixed for 20 min in methanol : acetone (1 : 1) at 220 uc. Cells were air dried for a few minutes at room temperature before incubation with the primary antibody, a mouse anti-hrv-16 mab diluted 1 : 3000 in PBS 22 /1 % BSA, for 45 min at 37 uc. After intensive washing with PBS 22, FITC-conjugated anti-mouse IgG antibody (Light Diagnostics) was added to the cells for 45 min at 37 uc in the dark. Cells were then washed three times with PBS 22 and stained with DAPI for 5 min at room temperature. After a final rinse with PBS 22, the coverslips were mounted in Fluoroprep mounting medium (biomérieux). Quantification of immunofluorescence-positive cells. Images were acquired with a Mirax microscope (Carl Zeiss) using a 620 objective. Image analysis was performed with Metamorph/ MetaXpress software (Molecular Devices). The blue channel recorded DAPI-stained nuclei, whereas the green channel recorded cells marked with the FITC-coupled antibody. The first step of processing was separation of the two channels. Respective channels were converted from 8 to 16 bits by multiplication. The CellScoring tool of the Metamorph software was applied to 16-bit versions of the blue and green channels. The parameters used to analyse the images were a cell minimal width of 13 mm and a maximal width of 40 mm, and an intensity above the local threshold of 20. Positive staining was determined in the cytoplasm (parameter Stained area ). The reported data included total cell number, positive-cell number and their relative percentages. Phylogenetic trees based on RNA sequences. RNA sequences for the 59 UTR, CL and IRES were aligned using the MUSCLE program (Edgar, 2004). Well-aligned regions were extracted using Gblocks (Castresana, 2000) and maximum-likelihood trees were estimated using PhyML (Guindon et al., 2010). One hundred bootstrap replicates were performed using the general time reversible (GTR) model for correcting nucleotide substitution rates. All trees were rooted on the simian virus 2 (SV2) outgroup strain. Phylogenetic trees based on RNA structure distances. Neighbour-joining trees were computed for RNA structures in the full 59 UTR, CL and IRES. The respective sequences were extracted for each strain used and folded into their minimum free-folding energy structure using RNAfold (Hofacker et al., 1994). Secondary RNA structures were converted to a tree representation of the structure, and a tree-edit distance for all pairwise combinations of strains was computed using RNA distance from the Vienna RNA package (Hofacker, 2009). Distance matrices were used with the program NEIGHBOUR from the PHYLIP package (Felsenstein, 2005) to compute the final neighbour-joining trees on RNA structure distances. Strains that group closely together in the trees show more similarity regarding the underlying RNA secondary structures, which can be interconverted with fewer edit operations than distantly placed strains. ACKNOWLEDGEMENTS We would like to thank Chantal Gaille, Carole Bampi, Manel Essaidi and Geneviève Mottet-Osman for technical assistance. We also thank Dominique Garcin and Laurent Roux for useful suggestions. This study was supported by the Swiss National Science Foundation (grant to C. T.; grant 32003B_ to L. K.) and partly supported by the Research Fund of the Department of Internal Medicine of the University Hospital and the Faculty of Medicine of Geneva; this Fund receives an unrestricted grant from AstraZeneca Switzerland, GlaxoSmithKline and Merck Sharp & Dohme. REFERENCES Arden, K. E., McErlean, P., Nissen, M. D., Sloots, T. P. & Mackay, I. M. (2006). Frequent detection of human rhinoviruses, paramyxoviruses, coronaviruses, and bocavirus during acute respiratory tract infections. J Med Virol 78, Borman, A. M., Deliat, F. G. & Kean, K. M. (1994). Sequences within the poliovirus internal ribosome entry segment control viral RNA synthesis. EMBO J 13, Castresana, J. (2000). Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol 17, Cordey, S., Gerlach, D., Junier, T., Zdobnov, E. M., Kaiser, L. & Tapparel, C. (2008). The cis-acting replication elements define human enterovirus and rhinovirus species. RNA 14, Cordey, S., Junier, T., Gerlach, D., Gobbini, F., Farinelli, L., Zdobnov, E. M., Winther, B., Tapparel, C. & Kaiser, L. (2010). Rhinovirus genome evolution during experimental human infection. PLoS ONE 5, e Domingo, E. & Holland, J. J. (1997). RNA virus mutations and fitness for survival. Annu Rev Microbiol 51, Du, Z., Yu, J., Ulyanov, N. B., Andino, R. & James, T. L. (2004). Solution structure of a consensus stem loop D RNA domain that plays important roles in regulating translation and replication in enteroviruses and rhinoviruses. Biochemistry 43, Edgar, R. C. (2004). MUSCLE: Multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32, Felsenstein, J. (2005). PHYLIP (phylogeny inference package) version 3.6. Distributed by the author. Department of Genome Sciences, University of Washington, Seattle. Garbino, J., Soccal, P. M., Aubert, J. D., Rochat, T., Meylan, P., Thomas, Y., Tapparel, C., Bridevaux, P. O. & Kaiser, L. (2009). Respiratory viruses in bronchoalveolar lavage: a hospital-based cohort study in adults. Thorax 64, Gromeier, M., Alexander, L. & Wimmer, E. (1996). Internal ribosomal entry site substitution eliminates neurovirulence in intergeneric poliovirus recombinants. Proc Natl Acad Sci U S A 93, Guindon, S., Dufayard, J. F., Lefort, V., Anisimova, M., Hordijk, W. & Gascuel, O. (2010). New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst Biol 59, Journal of General Virology 93

69 Experimental HRV and HEV recombination Hofacker, I. L. (2009). RNA secondary structure analysis using the Vienna RNA package. Curr Protoc Bioinform 26, Hofacker, L., Fontana, W., Stadler, F., Bonhoeffer, S., Tacker, M. & Schuster, P. (1994). Fast folding and comparison of RNA secondary structures. Monatsh Chem 125, Huang, T., Wang, W., Bessaud, M., Ren, P., Sheng, J., Yan, H., Zhang, J., Lin, X., Wang, Y. & other authors (2009). Evidence of recombination and genetic diversity in human rhinoviruses in children with acute respiratory infection. PLoS ONE 4, e6355. Karber, G. (1931). 50 % end-point calculation. Arch Exp Pathol Pharmakol 162, Kistler, A. L., Webster, D. R., Rouskin, S., Magrini, V., Credle, J. J., Schnurr, D. P., Boushey, H. A., Mardis, E. R., Li, H. & DeRisi, J. L. (2007). Genome-wide diversity and selective pressure in the human rhinovirus. Virol J 4, 40. Knowles, N. (2011). Enterovirus. enterovirus/enterovirus.htm. Livak, K. J. & Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2 2DDC t method. Methods 25, Lu, X., Holloway, B., Dare, R. K., Kuypers, J., Yagi, S., Williams, J. V., Hall, C. B. & Erdman, D. D. (2008). Real-time reverse transcription- PCR assay for comprehensive detection of human rhinoviruses. J Clin Microbiol 46, Lukashev, A. N. (2005). Role of recombination in evolution of enteroviruses. Rev Med Virol 15, Lukashev, A. N., Lashkevich, V. A., Ivanova, O. E., Koroleva, G. A., Hinkkanen, A. E. & Ilonen, J. (2005). Recombination in circulating Human enterovirus B: independent evolution of structural and nonstructural genome regions. J Gen Virol 86, Mäkelä, M. J., Puhakka, T., Ruuskanen, O., Leinonen, M., Saikku, P., Kimpimäki, M., Blomqvist, S., Hyypiä, T. & Arstila, P. (1998). Viruses and bacteria in the etiology of the common cold. J Clin Microbiol 36, McErlean, P., Shackelton, L. A., Lambert, S. B., Nissen, M. D., Sloots, T. P. & Mackay, I. M. (2007). Characterisation of a newly identified human rhinovirus, HRV-QPM, discovered in infants with bronchiolitis. J Clin Virol 39, McIntyre, C. L., McWilliam Leitch, E. C., Savolainen-Kopra, C., Hovi, T. & Simmonds, P. (2010). Analysis of genetic diversity and sites of recombination in human rhinovirus species C. J Virol 84, Palmenberg, A. C., Spiro, D., Kuzmickas, R., Wang, S., Djikeng, A., Rathe, J. A., Fraser-Liggett, C. M. & Liggett, S. B. (2009). Sequencing and analyses of all known human rhinovirus genomes reveal structure and evolution. Science 324, Palmenberg, A. C., Neubauer, D. & Skern, T. (2010). Genome organization and encoded proteins. In The Picornaviruses, pp Edited by E. Ehrenfeld, E. Domingo & R. P. Roos. Washington, DC: American Society for Microbiology. Perera, R., Daijogo, S., Walter, B. L., Nguyen, J. H. & Semler, B. L. (2007). Cellular protein modification by poliovirus: the two faces of poly(rc)-binding protein. J Virol 81, Racaniello, V. R. (2007). Picornaviridae: the viruses and their replication. Translation of the viral RNA. In Fields Virology, pp Edited by D. M. Knipe & P. M. Howley. Philadelphia, PA: Lippincott Williams & Wilkins. Rohll, J. B., Percy, N., Ley, R., Evans, D. J., Almond, J. W. & Barclay, W. S. (1994). The 59-untranslated regions of picornavirus RNAs contain independent functional domains essential for RNA replication and translation. J Virol 68, Ruohola, A., Waris, M., Allander, T., Ziegler, T., Heikkinen, T. & Ruuskanen, O. (2009). Viral etiology of common cold in children, Finland. Emerg Infect Dis 15, Santti, J., Hyypiä, T., Kinnunen, L. & Salminen, M. (1999). Evidence of recombination among enteroviruses. J Virol 73, Sawyer, M. H. (2001). Enterovirus infections: diagnosis and treatment. Curr Opin Pediatr 13, Simmonds, P. & Welch, J. (2006). Frequency and dynamics of recombination within different species of human enteroviruses. J Virol 80, Simmonds, P., McIntyre, C., Savolainen-Kopra, C., Tapparel, C., Mackay, I. M. & Hovi, T. (2010). Proposals for the classification of human rhinovirus species C into genotypically assigned types. J Gen Virol 91, Smura, T., Blomqvist, S., Paananen, A., Vuorinen, T., Sobotová, Z., Bubovica, V., Ivanova, O., Hovi, T. & Roivainen, M. (2007). Enterovirus surveillance reveals proposed new serotypes and provides new insight into enterovirus 59-untranslated region evolution. J Gen Virol 88, Tapparel, C., Junier, T., Gerlach, D., Cordey, S., Van Belle, S., Perrin, L., Zdobnov, E. M. & Kaiser, L. (2007). New complete genome sequences of human rhinoviruses shed light on their phylogeny and genomic features. BMC Genomics 8, 224. Tapparel, C., Cordey, S., Van Belle, S., Turin, L., Lee, W. M., Regamey, N., Meylan, P., Mühlemann, K., Gobbini, F. & Kaiser, L. (2009a). New molecular detection tools adapted to emerging rhinoviruses and enteroviruses. J Clin Microbiol 47, Tapparel, C., Junier, T., Gerlach, D., Van Belle, S., Turin, L., Cordey, S., Mühlemann, K., Regamey, N., Aubert, J. D. & other authors (2009b). New respiratory enterovirus and recombinant rhinoviruses among circulating picornaviruses. Emerg Infect Dis 15, Todd, S., Towner, J. S. & Semler, B. L. (1997). Translation and replication properties of the human rhinovirus genome in vivo and in vitro. Virology 229, Yozwiak, N. L., Skewes-Cox, P., Gordon, A., Saborio, S., Kuan, G., Balmaseda, A., Ganem, D., Harris, E. & DeRisi, J. L. (2010). Human enterovirus 109: a novel interspecies recombinant enterovirus isolated from a case of acute pediatric respiratory illness in Nicaragua. J Virol 84,

70 P1-2A region exchange between members of the same HRV species is viable, whereas P1-2A interspecies recombinants are not functional. Experimental design. Four chimeric HRV genomes were constructed in which the P1 region coding for the four capsid proteins was replaced by the corresponding part of the genome derived from another HRV, either from the same species or from a different species. As the cleavage between VP1 and 2A is performed by the 2A protease, the P1 region together with 2A was switched, to avoid cleavage site incompatibilities (Figure 14). Polyprotein cleavage sites of HRV and HEV genomes used in this thesis are shown in Table 3. Nucleotide sequence homologies between P1-2A regions from the different HRVs used to create these chimeric genomes are shown in Table 4. Each chimeric derivative was in vitro transcribed and transfected in Hela Ohio cells. Viral multiplication was assessed by the presence of a cytopathic effect (CPE). If no CPE was visible seven days after transfection, an immunofluorescence (IF) assay using mab J2 was performed to confirm the absence of viral replication. 56

71 Figure 14. Chimeric HRV genomes in which the P1 region (indicated by the curly brace) together with the 2A region is replaced by homologous sequences from a different HRV genome. Construct names are indicated on the left and HRV types from which genomes are derived are indicated by different shades of grey. Table 3. Polyprotein cleavage sites of HRV and HEV genomes used in recombination experiments. Polyprotein cleavage site Protease HRV-A16 HRV-A39 HRV-A81 HRV-B14 HRV-C11 HEV-A71 VP4-VP2 - Q-S Q-S Q-S N-S M-S K-S VP2-VP3 3C Q-G Q-G Q-G Q-G Q-G Q-G VP3-VP1 3C Q-N Q-N Q-N E-G Q-N Q-G VP1-2A 2A V-G A-G V-G Y-G A-G L-G 2A-2B 3C Q-G Q-G Q-G Q-G Q-G Q-G 2B-2C 3C E-S E-S E-S Q-A Q-S Q-S 2C-3A 3C Q-G Q-G Q-G Q-G Q-G Q-G 3A-3B 3C Q-G Q-G Q-G Q-G Q-G Q-G 3B-3C 3C Q-G Q-G Q-G Q-G Q-G Q-G 3C-3D 3C Q-G Q-G Q-G Q-G Q-G Q-G 57

72 Table 4. Nucleotide sequence homologies in percents between the different P1-2A regions used in this study. HRV-A16 HRV-A81 HRV-C11 HRV-B HRV-A HRV-A Replication ability of the P1-2A intra- and interspecies chimeras. Transfection of the HRV-A16/HRV-A81 P1-2A construct yielded infectious viral particles that could be passaged successfully. Sequencing confirmed the chimeric nature of the replicating virus. Transfection of the interspecies HRV-A16/HRV-14B P12A, HRV- A16/HRV-C11 and HRV-C11/HRV-A16 P1-2A chimeras did not result in viral replication (Figure 15). 58

73 B. Figure 15. Immunofluorescence using the mab J2 antibody performed 16 hours post transfection with the indicated P1-2A chimeras (panels A, C and D), with a non chimeric HRV-B14 positive control (panel B); the negative control (untransfected) is shown in panel E. HRV-A/HRV-A P1-2A infected cells are indicated by arrow. 59

74 Non replicative recombination ability of artificially engineered defective viral genomes Experimental design HRV genomes deleted either in the 5 UTR, including RNA structures essential for translation and replication, or in the 3 end of the viral genome including the 3D region coding for the viral RNA polymerase were generated. These altered genomes were theoretically unable to be translated and replicated since the 5 UTR contains RNA structures crucial for both processes and the 3D polymerase is essential for viral replication. Two different HRV-A16 genomes deleted in the 5 UTR were engineered. The first, HRV-A16 del5 UTR ( ) (Figure 16), lacks the 3 half of the 5 UTR, including IRES domains IV and V. The second deleted HRV- A16 construct, HRV-A16 del5 UTR (1-434) (Figure 16), lacks the first 434 nucleotides of the 5 UTR, including a major part of the IRES and the CL, essential for replication. However, it retains the 3 end of the 5 UTR (191 nucleotides), known to be an HRV recombination site. Co-transfection of 5 and 3 deletion constructs with complementary genomes in cell cultures could theoretically take advantage of non replicative RNA recombination to generate complete and functional genomes. The method was first tested with 5 and 3 deleted HRV- A16 genomes (Figure 18A and 18B). Then, a 5 UTR-deleted HRV-A16 genome was cotransfected with a 3 deleted HRV-A39 genome (HRV-A39 del3 ( ), Figure 16), lacking a major mart of the 3D region sequence coding for the viral RNA polymerase, to study intraspecies non replicative RNA recombination (Figure 18C). These two HRV genomes share a 75.8% nucleotide sequence homology. Finally, HRV-A16 del5 UTR (1-434) was cotransfected with a 3 deleted HRV-C11 genome (HRV-C11 del3 ( ), Figure 16) to study interspecies non replicative RNA recombination (Figure 18D). HRV-A16 and HRV-C11 60

75 share a 58.3% nucleotide sequence homology. Upon appearance of a CPE, resulting virus was passaged to confirm the infectivity, before being sequenced to map recombination sites. Again, if no CPE was visible seven days after transfection, an immunofluorescence (IF) assay with a mab J2 was performed to confirm the absence of viral particle production. Figure 16. Schematic representation of the deleted HRV genomes used for non replicative RNA recombination experiments. The crosses indicate the regions deleted for each construct. Construct names are indicated on the left. The SfoI restriction site used as a 5 sequence marker is indicated by white arrows heads. The four nucleotide substitutions used as a 3 sequence marker in the HRV-A16 del5 UTR ( ) 3 marker genome are indicated by the grey arrow head. The five nucleotide substitutions used as a 3 sequence marker in the HRV-A16 del5 UTR (1-434) 3 marker genome are indicated by the black arrow head. 61

76 Non replicative RNA recombination occurs with high frequency between two viruses of the same serotype To confirm the replication inability of the 5 and 3 deleted genomes, each of them was transfected individually. Then, as a proof of principle, HRV-A16 del5 UTR ( ) and HRV- A16 del3 ( ), were co-transfected in Hela Ohio cells (Figure 18A). CPE occurred in three out of four co-transfection experiments. Furthermore, upon co-transfection of HRV-A16 del5 UTR (1-434) 3 marker with HRV-A16 del3 ( ) (Figure 18B), CPE occurred in four out of four transfections. Infected cell Supernatants were passaged on fresh cells, which reproduced a CPE, indicating the presence of infectious virus. Sequencing of the 5 and 3 ends of the HRV-A16 genome confirmed the recombinant nature of these viruses. The 5- UTR-ORF junction was devoid of the SfoI site, and the 3 UTR harboured the expected 3 UTR marker sequence (Figure 18A and B). Recombination sites could not be mapped since the sequence of the two parents was identical except for the markers. 62

77 Non replicative RNA recombination occurs with low frequency between two viruses of the same species Co-transfection of QCHRV-A16 del5 UTR and HRV-A39 del3d RNAs (Figure 18C) resulted in replicating virus in two out of eight transfected wells. Infectivity of these two viruses was confirmed by passageing. Their recombinant nature was documented by sequencing, and their respective recombination site was mapped (Figure 18C). One recombination site was located at position 5893 and the other at 5900 of the HRV-A39 genome, corresponding to the beginning of the 3D gene. Ten subsequent transfections gave rise to two additional HRV-A39/HRV-A16 recombinants, whose recombination site was found to be located at position 880 and 2393 of the HRV-A39 genome, corresponding tovp2 and VP1 regions, respectively. Non replicative RNA recombination could not be highlighted between viruses of different species Co-transfection of HRV-A16 del5 UTR (1-434) and HRV-C11 del3 ( ) (Figure 18D) did not cause any CPE, even after 40 co-transfection experiments. However, IF performed on 11 co-transfection experiments seven days after co-transfection revealed a cluster of a few infected cells in each of four out of 11 wells (Figure 17). Passageing of the corresponding supernatant did not result in CPE suggesting that the viruses, although being able to infect neighboring cells, were not fit enough to multiply sufficiently to be passaged. 63

78 Figure 17. Immunofluorescence using the mab J2 antibody seven days post co-transfection of QCHRV-A16 del5 UTR (1-434) and HRV-C11 del3 ( ) RNAs (panel A) and of HRV-A16 del5 UTR (1-434) and HRV-A39 del3 ( ) RNAs (panel B). Figure 18. HRV Recombinants rescued from co-transfection experiments. HRV-A16/HRV-A16 recombinants devoid of the 5 SfoI restriction site marker and harbouring the 3 marker sequence (A) and (B). HRV-A39/HRV- A16 recombinants with recombination sites indicated by small arrows (C). HRV-A39 sequences are represented in light gray and HRV-A16 sequences in white. Recombinants resulting from HRV-A16 del5 UTR(1-434) and HRV-C11 del3 ( ), shown in dark grey, could not be passaged and sequenced (D). 64

79 RHINOVIRUS RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS: ARTICLE 2 Critical Analysis of Rhinovirus RNA Load Quantification by Real-Time Reverse Transcription-PCR. Schibler M, Yerly S, Vieille G, Docquier M, Turin L, Kaiser L, Tapparel C. J Clin Microbiol Sep;50(9):

80 SUPPLEMENTAL MATERIAL REFERENCES CONTENT ALERTS Critical Analysis of Rhinovirus RNA Load Quantification by Real-Time Reverse Transcription-PCR Manuel Schibler, Sabine Yerly, Gaël Vieille, Mylène Docquier, Lara Turin, Laurent Kaiser and Caroline Tapparel J. Clin. Microbiol. 2012, 50(9):2868. DOI: /JCM Published Ahead of Print 20 June Updated information and services can be found at: These include: Supplemental material This article cites 24 articles, 11 of which can be accessed free at: Receive: RSS Feeds, etocs, free alerts (when new articles cite this article), more» Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve Information about commercial reprint orders: To subscribe to to another ASM Journal go to:

81 Critical Analysis of Rhinovirus RNA Load Quantification by Real- Time Reverse Transcription-PCR Manuel Schibler, a Sabine Yerly, a Gaël Vieille, a Mylène Docquier, b Lara Turin, a Laurent Kaiser, a and Caroline Tapparel a Laboratory of Virology, Division of Infectious Diseases and Division of Laboratory Medicine, University of Geneva Hospitals, Geneva, Switzerland, a and Genomics Platform, NCCR Frontiers in Genetics, University of Geneva, Geneva, Switzerland b Rhinoviruses are the most frequent cause of human respiratory infections, and quantitative rhinovirus diagnostic tools are needed for clinical investigations. Although results obtained by real-time reverse-transcription PCR (RT-PCR) assays are frequently converted to viral RNA loads, this presents several limitations regarding accurate virus RNA quantification, particularly given the need to reliably quantify all known rhinovirus genotypes with a single assay. Using an internal extraction control and serial dilutions of an in vitro-transcribed rhinovirus RNA reference standard, we validated a quantitative one-step real-time PCR assay. We then used chimeric rhinovirus genomes with 5=-untranslated regions (5=UTRs) originating from the three rhinovirus species and from one enterovirus to estimate the impact of the 5=UTR diversity. Respiratory specimens from infected patients were then also analyzed. The assay quantification ability ranged from 4.10 to 9.10 log RNA copies/ml, with an estimated error margin of 10%. This variation was mainly linked to target variability and interassay variability. Taken together, our results indicate that our assay can reliably estimate rhinovirus RNA load, provided that the appropriate error margin is used. In contrast, due to the lack of a universal rhinovirus RNA standard and the variability related to sample collection procedures, accurate absolute rhinovirus RNA quantification in respiratory specimens is currently hardly feasible. Human rhinoviruses (HRVs) are small nonenveloped viruses containing a positive-strand RNA genome. They belong to the Enterovirus genus, which is part of the Picornaviridae family. HRVs are important human pathogens and the most frequent cause of respiratory infections. Their tropism is not restricted to the upper respiratory tract, and they can cause complications in the lower respiratory tract, especially in children, the elderly, and immunocompromised patients (5 7, 15, 16). Many HRVs, in particular those belonging to the newly discovered HRV-C species, do not grow under traditional cell culture conditions. Therefore, molecular diagnostic tools, such as real-time reverse transcription- PCR (RT-PCR), are the methods of choice for diagnosing HRVs. Thanks to numerous clinical studies, data have been gathered regarding implication of HRV in the exacerbation of underlying diseases, such as cystic fibrosis, asthma, and chronic obstructive pulmonary disease (1, 3, 7, 8, 10, 17, 18, 20, 21, 25), as well as concerning the possible impact of a given HRV species on disease severity (9, 11 13, 23). However, there are inconsistent data about whether viral load is correlated with disease severity, and the threshold cycle (C T ) values from real-time RT-PCR are converted too frequently into viral copies per milliliter without any validation. This is particularly important because many published assays include primers with degenerate nucleotides or probe sequences that are biased toward a given genotype or species (4, 14, 22, 24). We validated previously a two-step real-time RT-PCR assay, named Panenterhino/Ge/08, which could detect all known HRV genotypes and, to a lesser extent, human respiratory enteroviruses (HEVs) (22). In this study, we first validated the Panenterhino/ Ge/08 assay in a one-step quantitative format using an internal extraction control and serial dilutions of an in vitro-transcribed rhinovirus RNA reference standard. HRV-positive clinical specimens were then quantified with this assay, and the results were compared to those obtained with the two-step Panenterhino/ Ge/08 real-time RT-PCR. Interassay reproducibility was analyzed, and the level of imprecision linked to the use of a single realtime RT-PCR assay to detect the whole HRV group consisting of genetically variable types was experimentally investigated. Other sources of inaccuracy and imprecision occurring in the HRV RNA quantification process are discussed. MATERIALS AND METHODS Plasmids. The plasmids QCpR16.11 and pwr3.26-hrv-14 encode the full-length genomes of the HRV-A16 and the HRV-B14 strains, respectively, and were kindly provided by W.-M. Lee (University of Wisconsin). The HRV-Ca/A16, HRV-B37/A16, and HEV-A71/A16 plasmids encode chimeric viral genomes in which the 5=-untranslated region (5=UTR) of the HRV-A16 is replaced with the 5=UTR of an HRV-Ca, HRV-B37, or HEV-A71 strain, respectively, and were constructed as previously described (19). Production and quantification of in vitro RNA transcripts. For each construct, 5 to 10 g of plasmid was linearized at a unique SacI restriction site downstream of the 3=-viral poly(a) tail. The MEGAscript T7 kit (Ambion) was used to synthesize RNA transcripts from the linear templates for 3hat37 C. The transcripts were then purified with the RNeasy minikit (Qiagen). In vitro-transcribed RNAs were quantified by measuring the optical density (see below), diluted to 1 g/ l, and checked by 0.1% sodium dodecyl sulfate 1% agarose gel analysis. The transcripts were diluted into the NucliSens easymag extraction buffer 3 (bio- Mérieux). The weight of one RNA molecule in grams was determined using the mean nucleotide molecular weight and the number of nucleotides per RNA molecule. Received 20 December 2011 Returned for modification 28 January 2012 Accepted 5 June 2012 Published ahead of print 20 June 2012 Address correspondence to Manuel Schibler, manuel.schibler@hcuge.ch. Supplemental material for this article may be found at Copyright 2012, American Society for Microbiology. All Rights Reserved. doi: /jcm Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve 2868 jcm.asm.org Journal of Clinical Microbiology p September 2012 Volume 50 Number 9

82 Rhinovirus RNA Quantification by Real-Time RT-PCR Patient cohorts and clinical specimens. Nasopharyngeal swab (NPS) and bronchoalveolar lavage (BAL) specimens were obtained from patients enrolled in a cohort of lung transplant recipients (September 2008 to November 2010). These respiratory specimens were screened using the Panenterhino/Ge/08 two-step assay, and stored positive samples were reanalyzed for viral load quantification. This study was approved by the institutional review board and the ethics committee of the University of Geneva Hospitals. Written informed consent was obtained from all individuals. To test the linearity of the assay in clinical samples, an NPS specimen and a BAL specimen with high HRV RNA load were selected; they were diluted 10-fold into a pool of HRV-negative NPS and BAL specimens, respectively, and into Universal transport medium (UTM; Copan). Primers and hydrolysis probes. The two-step Panenterhino/Ge/08 and canine distemper virus (CDV) real-time RT-PCR assays were performed as previously described (2, 22). For CDV one-step real-time RT- PCRs, the primers and hydrolysis probe were used at final concentrations of 0.9 M and 0.5 M, respectively. The HRV-A16-specific two-step realtime RT-PCR assay (HRV-A16 3D), designed to amplify nucleotides (nt) 6903 to 6970 in the 3D gene of the HRV-A16 genome (GenBank accession no. L24917), was used as previously described (19). The human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) endogenous control (VIC/TAMRA [6-carboxytetramethylrhodamine] probe; Primer limited) assay (Applied Biosystems) and the CELL control r-gene kit (Argene) were used according to the manufacturer s instructions. Available sequences of primers and hydrolysis probes are listed in Table S1 in the supplemental material. RNA extraction and real-time RT-PCR. As an internal control, 10 l of a homogeneous dilution of a CDV stock was added to each sample before extraction. The RNA was extracted from clinical samples (190 lof clinical sample) using the NucliSens easymag magnetic bead system (biomérieux) and eluted into 25 l. Five microliters of the extracted RNA was reverse transcribed with Superscript II (Invitrogen) with random hexamers (Roche) according to the manufacturer s instructions in a reaction volume of 20 l. Forty microliters of water was added to the RT product to have enough volume to conduct all assays in parallel. Five microliters of diluted cdna was used for each real-time RT-PCR. For the one-step reactions, the extracted RNA was diluted so as to obtain the same copy number in each real-time RT-PCR format. For the real-time PCRs, the chimeric plasmids were analyzed with the HRV-A16 3D and Panenterhino/Ge/08 assays, and the reverse-transcribed cdnas were analyzed with the Panenterhino/Ge/08 and the CDV assays using the TaqMan Universal PCR master mix (Applied Biosystems) in a 7500 or a 7000 Applied Biosystems thermocycler. The following cycling conditions were used: 50 C for 2 min, 95 C for 10 min, and 45 cycles of 95 C for 15 s and 60 C for 1 min. For one-step real-time PCRs, RNAs were analyzed with the Panenterhino/Ge/08 and CDV assays using the QuantiTect probe RT- PCR kit (Qiagen) according to the manufacturer s instructions in a 7000 Applied Biosystems thermocycler. The following cycling conditions were used: 50 C for 30 min, 95 C for 15 min, and 45 cycles of 94 C for 15 s and 60 C for 1 min. Results were analyzed using the program SDS version 1.4 (Applied Biosystems). Standard curves and RNA quantification. To establish a standard curve for the quantitative real-time RT-PCR experiments, we used serial 10-fold dilutions (from to 62.5 copies/reaction) of an in vitrotranscribed full-length HRV-A16 RNA that was quantified by optical density. To check the accuracy of RNA quantification by optic density measurements, the amount of in vitro-transcribed RNA was measured in triplicate at two different time points separated by a freeze-thaw cycle. The mean values were 5.77 ng/ l (standard deviation [SD], 0.15 ng/ l) for day 1 and 5.73 ng/ l (SD, 0.31 ng/ l) for day 2 (corresponding to RNA copies/ml and copies/ml, respectively). To evaluate the linearity of the measurements, the RNA stock was diluted 10-fold within the range of quantification by optical density and measured in triplicate. The first dilution had a mean concentration of ng/ l (SD, ng/ l), the second ng/ l (SD, 0.17 ng/ l), and the third 4.67 ng/ l (SD, 0.51 ng/ l), which correspond to , , and RNA copies/ml, respectively. The RNA of each respiratory specimen used for HRV RNA load quantification was extracted twice at two different time points. Each set of extracted RNA was tested in duplicate with the one-step and two-step Panenterhino/Ge/08 assays and with the internal control CDV assay. The duplicates from the Panenterhino/Ge/08 assay were normalized as follows: the lowest CDV C T value of each run was set as the reference value, and the difference in the CDV C T values between the sample and the reference was subtracted from the value obtained from the Panenterhino/ Ge/08 C T assay. The means of the normalized C T values from the real-time RT-PCR duplicates were used to quantify the RNA copy numbers per reaction using the slope-intercept form. The number of RNA copies per reaction was then corrected for the dilution factor performed between the extraction and the real-time RT-PCR ( fold) to obtain a final viral RNA copy number per ml of initial sample (described above). PCR efficiency (E) was calculated using the formula E 10 1/slope 1. Statistical analysis. To evaluate the statistical significance of differences in RNA quantification values obtained with the one-step and twostep Panenterhino/Ge/08 assays, P values were determined using the paired t test. RESULTS Analytical sensitivity and linear range of the one-step Panenterhino/Ge/08 real-time RT-PCR assay. The limit of detection (LOD) and the limit of quantification (LOQ) of the one-step Panenterhino/Ge/08 assay were assessed and compared to those obtained with the two-step assay using three 10-fold dilution series of an HRV-A16 RNA transcript. The one-step and two-step reactions were performed in parallel on each diluted RNA. The LOD were 3.10 log copies/ml for the one-step assay and 2.10 log copies/ml for the two-step assay. For both assays, the LOQ was 4.10 log copies/ml, and linearity was conserved from 4.10 to 9.10 log copies/ml. Intra-assay reproducibility was better for the onestep assay (r ) than for the two-step assay (r ) (Fig. 1A and B). The linearity of the one-step Panenterhino/Ge/08 assay on clinical specimens was also analyzed using 10-fold dilution series of a Panenterhino/Ge/08-positive NPS and a Panenterhino/Ge/ 08-positive BAL specimen. The assay was linear for clinical specimens with C T values between 22 and 34 (r for the NPS experiment and r for the BAL experiment). However, the PCR efficiency (E) was lower with these clinical specimens (E 80.3 to 80.7%) than with the in vitro-transcribed RNA (E 84.9 to 99.1%). Comparable results were obtained when the HRV-positive specimens were diluted into the widely used UTM (data not shown). Sources of HRV RNA load quantification bias in clinical specimens. (i) Quantification with one-step versus two-step real-time RT-PCR. The HRV viral RNA load was determined in a collection of 44 NPS and 6 BAL picornavirus-positive specimens using the one- and two-step Panenterhino/Ge/08 assays. The sequences of 22 of the 50 picornavirus-positive specimens were available: 12 were HRV-A members, 1 was an HRV-B member, and 9 were HRV-C members. In each real-time RT-PCR experiment, a four-dilution series of a reference RNA standard of known quantity was included. The RNA load was calculated as described in Materials and Methods. Samples presenting less than 4.10 log Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve September 2012 Volume 50 Number 9 jcm.asm.org 2869

83 Schibler et al. FIG 1 Linearity and variability of the Panenterhino/Ge/08 two-step (A) and one-step (B) real-time RT-PCR on three 10-fold dilution series of in vitrotranscribed HRV-A16 RNA (A and B). The standard curves, slopes, and r 2 values are shown. copies/ml, and thus below the LOQ, were not included in the quantitative analysis. The comparison between one-step and two-step Panenterhino/ Ge/08 HRV RNA quantifications is illustrated by the Bland- Altman plot presented in Fig. 2, and the raw data are given in Table S2 in the supplemental material. The median viral RNA loads were 5.68 log RNA copies/ml (range, 4.44 to 7.45 log RNA copies/ml) with the one-step assay and 5.83 log RNA copies/ml (range, 4.57 to 7.53 log RNA copies/ml) with the two-step assay. On average, viral loads obtained with the two-step assay were 0.15 log RNA copy/ml (SD, 0.47 log RNA copy/ml) higher than those obtained using the one-step assay (P 0.026, paired t test). Differences above 0.5 log RNA copy/ml were observed in 36% of the samples and within all viral RNA levels. The mean percent coefficient of variation (% CV) related to one-step versus two-step HRV RNA quantification was 4.9% (range, 0.3 to 13.0%). (ii) Interassay variability. Interassay quantification reproducibility was evaluated using 12 HRV-positive respiratory samples with viral loads ranging from 4.4 to 7.6 log RNA copies/ml and processed by two different laboratory technicians. For each specimen, two biological replicates were performed from RNA extraction to one-step or two-step real-time RT-PCR. The mean % CV of the log RNA copies/ml was 2.1% (range, 0.3 to 3.8%) for the one-step and 6.8% (range, 1.3 to 14.0% for the two-step assay) (see Fig. 4 and see Table S3 in the supplemental material). FIG 2 One-step Panenterhino/Ge/08 HRV RNA quantification values compared to two-step Panenterhino/Ge/08 HRV RNA quantification values. A Bland-Altman plot shows the distribution of HRV RNA quantification values in log copies/ml, compared to the values obtained with the two-step assay. Two-step HRV quantification values are represented on the x axis. Differences in quantification values between the one-step and two-step assays in log copies/ml are represented on the y axis. The dashed line represents the mean difference of log copies/ml between both assays (two-step assay one-step assay 0.15 log), and the dotted lines represent the mean difference of log copies/ml between both assays 2 SD (1.09) and 2SD( 0.79), respectively. Raw data are available in Table S2 in the supplemental material. (iii) Sample heterogeneity. For the same 12 samples, the GAPDH endogenous transcript control assay and the CELL control r-gene assay, which amplify, respectively, the mrna and DNA of the host cells, were run in parallel (see Table S3 in the supplemental material). The results revealed an important variability in cell content among the different clinical specimens, with GAPDH C T values ranging from to undetected and CELL control r-gene C T values ranging from to There was no correlation between HRV RNA load and the amount of human cells in these specimens. Indeed, some samples enriched in cells presented a low viral load, and, inversely, samples with low cellular RNA or DNA presented a high viral load. (iv) Target genetic variability. To experimentally evaluate the variations in quantification related to the genetic variability of HRVs, we took advantage of plasmids that encode chimeric viral genomes (19). These genomes have a common HRV-A16 polyprotein sequence, but the 5=UTRs are derived from three different HRV species and one HEV-A species (HRV-B37, HRV-A16, HRV-C11, and HEV-A71, respectively). Ten-fold dilution series were performed for each chimera, and for each dilution, the number of plasmid copies per ml was quantified with the Panenterhino/Ge/08 assay using the standard curve derived from the CT values obtained with the specific assay (HRV- A16 3D) run on the same template. Figure 3 illustrates the viral genome log copies/ml obtained using the four different constructs. In comparison to HRV-A16 3D, the mean viral genome loads were as follows: on the HRV-Ca construct, 0.38 log copies/ml (95% confidence interval [CI], 0.17 to 0.59); on the HRV-A16 construct, 0.23 log copies/ml (95% CI, 0.08 to 0.37); on the HRV-B37 construct, 0.17 log copies/ml (95% CI, 0.33 to 0.02); and on the HEV A-71 construct, 0.77 log copies/ml (95% CI, 0.47 to 1.07). Of note, differences above 0.5 log copies/ml were only observed for the HEV A-71/A16 construct. The mean % CV between values obtained with 5 10-fold dilutions of each of the different constructs and HRV-A16 3D was Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve 2870 jcm.asm.org Journal of Clinical Microbiology

84 Rhinovirus RNA Quantification by Real-Time RT-PCR FIG 3 Experimental evaluation of HRV nucleic acid quantification variation linked to HRV genetic variability. The plot shows chimeric plasmid quantifications expressed in log copies/ml using the Panenterhino/Ge/08 assay (y axis) compared to the HRV-A16 3D values (x axis), using five 10-fold dilutions of chimeric plasmids, ranging from 4.30 to 8.30 copies/ml. Raw data are available in Table S4 in the supplemental material. then calculated. When considering only the three HRV 5=UTRs, the mean % CV was 5.2% (range, 2.6 to 7.7%), whereas it was 7.0% (range, 5.6 to 8.4%) when the HEV 5=UTR was included (Fig. 4). Figure 4 summarizes the sources of viral load quantification errors analyzed in this study. The % CV linked to one-step or two-step interassay variability (see Table S3 in the supplemental material), to HRV/HRV and HRV/HEV target genetic variability (Fig. 3; see Table S4 in the supplemental material), and to the assay used to quantify viral load in respiratory specimens (Fig. 2; see Table S2 in the supplemental material) are shown. Of note, when combining the % CV linked to interassay variability and that related to target genetic variability, the mean % CV of the HRV log RNA copies/ml ranged between 7.3% and 9.1% for the one-step assay and between 12.1% and 13.8% for the two-step assay. DISCUSSION Molecular techniques that detect RNA viruses have quickly evolved, and real-time RT-PCR provides the opportunity to quantify viral loads. The one-step real-time RT-PCR is a welcome simplification of the two-step RT-PCR, saves time, and results in a lower risk of technical errors. In this study, we adapted a two-step real-time RT-PCR assay, designed on the basis of an alignment of all HRV-A and HRV-B genotype sequences, as well as including 11 divergent HRV-C 5=UTR sequences (22) for one-step use. We have carefully assessed the feasibility and the limitations of its application for viral RNA load quantification in clinical specimens. Comparison of one-step and two-step formats on serially diluted HRV RNA samples revealed that both assays were linear between and RNA copies/ml and had an LOD of RNA copies/ml for the one-step assay and RNA copies/ml for the two-step assay. However, the reproducibility of the one-step assay was higher, implying that the quantification reliability is better in this format. Application of the one-step and two-step assays for the quantification of HRV load among clinical specimens, as well as comparison of experiments performed by two different laboratory technicians, supported the increased reliability of the one-step assay. FIG 4 Putative sources of real-time RT-PCR viral load quantification variation. Box plots show the variation of viral load quantification (% CV) related to interassay variability (one-step and two-step assays), HRV genetic variability, HRV/HEV genetic variability, and one-step assay versus two-step assay variability. In summary, the one-step Panenterhino/Ge/08 real-time RT-PCR assay is recommended for diagnostic use and for HRV RNA quantification. Several factors can lead to inaccurate measurements and limit the quantification of HRV RNAs in respiratory samples. Therefore, the viral loads in this study, as well as in other clinical studies, have to be considered with caution. The most important limitation is probably the lack of an accurately quantified international reference standard RNA used for the establishment of standard curves, which makes absolute quantification currently impossible. However, in vitro-transcribed HRV-A16 RNA measurement by optical densitometry was revealed to be reproducible and quantitative during this study, enabling precise quantification, which allows comparison of viral loads in different specimens. As the PCR efficiency may vary from one experiment to another, it is important to include this standard RNA in every single experiment to obtain a standard curve that reflects the PCR efficiency of each experiment. Still, as demonstrated by the serial dilutions of the clinical specimens, the PCR efficiency of the standard will not be totally comparable to the efficiency of the same PCR performed on clinical samples. We also demonstrated that HRV RNA quantification values differed significantly between the one-step and the two-step technologies. It is therefore important to use the same format when comparing such quantification results. The technical reproducibility may also have some limitations. In our study, the interassay reproducibility was satisfactory with the one-step format, as shown by the low % CV. In contrast, the interassay reproducibility was much lower with the two-step assay. Additionally, the use of a single real-time RT-PCR assay to detect over 150 described HRV genotypes has trade-offs. Although the Panenterhino/Ge/08 assay targets a conserved sequence stretch, the primers and probe contain degenerate positions, and there is a possibility of mismatches because of genetic variability. Moreover, two forward primers are present in the mix, Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve September 2012 Volume 50 Number 9 jcm.asm.org 2871

85 Schibler et al. one of which was designed to detect the more divergent HRV-Cs (22). Therefore, the efficiency of the RT-PCR will vary, depending on the genotype being amplified. This bias is important and needs to be taken into account not only for the absolute, but also for the relative quantification of HRVs. In this study, we estimated the impact of the 5=UTR genetic variability on viral quantification by the use of chimeric 5=UTR/polyprotein plasmids. As expected, HEV-A71, which has the most divergent target sequence with respect to our assay, generated the highest differences. Based on our experiments, we estimate that the variation in quantification related to HRV and HEV genetic variability is less than 10%. When combining the % CV linked to interassay variability and that related to target genetic variability in order to estimate global variability related to HRV RNA quantification in respiratory specimens, the mean % CV of the HRV log RNA copies/ml remained below 10% for the one-step assay, whereas it was around 14% with the two-step assay. It should be noted, however, that interassay variability may depend on the target sequence being amplified and that interassay variability may thus not be totally independent from target variability. Finally, and perhaps most importantly, the nature of the respiratory specimens renders the interpretation of RNA quantifications difficult. Sampling techniques between individuals may differ considerably, especially regarding the NPS. Depending on which part of the nasal cavity is reached and how much force is applied to the mucosa, the quantities of viral RNA collected may vary significantly. Of note, we have shown that there was no strict correlation between the HRV RNA load and the amount of human cells in the specimens tested in our study. This may be explained by the fact that HRV viral RNA is not exclusively found inside the cells. However, the presence of a few cells in a specimen may indicate poor sampling technique, and the resulting determination of the viral load will thus likely be underestimated. Hence, when comparing the viral loads between two different specimens, it might be advisable to take human cell numbers into account to assess the reproducibility of the sampling procedure. Importantly, because the sampling techniques used for NPS and BAL specimens are completely different, the viral RNA quantities present in these two different kinds of samples should not be compared. In summary, we validated the Panenterhino/Ge/08 real-time RT-PCR assay for one-step use and for HRV RNA quantification. Several technical difficulties limit the use of real-time RT-PCR for the quantification of the absolute viral load in respiratory specimens, and the lack of an accurately quantified international reference RNA for the establishment of standard curves is probably the most important limitation. The variability related to sampling procedures, which complicates the interpretation of HRV RNA quantification results, may be assessed by the use of an internal cellularity control. Despite these limitations, one-step real-time RT-PCR HRV RNA quantification was revealed to be precise, with an estimated global variability below 10%. It therefore enables comparison of HRV RNA amounts in respiratory specimens, provided that the samples are collected using a standard procedure and that the same reference RNA standard is used. ACKNOWLEDGMENTS This study was supported by the Swiss National Science Foundation (grant no to C.T. and grant no B_ to L.K.) and partly supported by the Research Fund of the Department of Internal Medicine of the University Hospital and the Faculty of Medicine of Geneva. This fund receives an unrestricted grant from AstraZeneca Switzerland, GlaxoSmithKline, and Merck Sharp & Dohme. REFERENCES 1. Burns JL, et al Respiratory viruses in children with cystic fibrosis: viral detection and clinical findings. Influenza Other Respir. Viruses 6: Cordey S, et al Rhinovirus genome evolution during experimental human infection. PLoS One 5:e doi: /journal.pone de Almeida MB, et al Rhinovirus C and respiratory exacerbations in children with cystic fibrosis. Emerg. Infect. Dis. 16: Do DH, et al A one-step, real-time PCR assay for rapid detection of rhinovirus. J. Mol. Diagn. 12: Garbino J, et al Lower respiratory viral illnesses: improved diagnosis by molecular methods and clinical impact. Am. J. Respir. Crit. Care Med. 170: Garbino J, et al Respiratory viruses in bronchoalveolar lavage: a hospital-based cohort study in adults. Thorax 64: Hayden FG Rhinovirus and the lower respiratory tract. Rev. Med. Virol. 14: Heymann PW, Platts-Mills TA, Johnston SL Role of viral infections, atopy and antiviral immunity in the etiology of wheezing exacerbations among children and young adults. Pediatr. Infect. Dis. J. 24:S217 S Iwane MK, et al Human rhinovirus species associated with hospitalizations for acute respiratory illness in young US children. J. Infect. Dis. 204: Jackson DJ, Johnston SL The role of viruses in acute exacerbations of asthma. J. Allergy Clin. Immunol. 125: Jin Y, et al Prevalence and clinical characterization of a newly identified human rhinovirus C species in children with acute respiratory tract infections. J. Clin. Microbiol. 47: Kiang D, et al Molecular characterization of a variant rhinovirus from an outbreak associated with uncommonly high mortality. J. Clin. Virol. 38: Lau SK, et al Clinical features and complete genome characterization of a distinct human rhinovirus (HRV) genetic cluster, probably representing a previously undetected HRV species, HRV-C, associated with acute respiratory illness in children. J. Clin. Microbiol. 45: Lu X, et al Real-time reverse transcription-pcr assay for comprehensive detection of human rhinoviruses. J. Clin. Microbiol. 46: Papadopoulos NG, et al Rhinoviruses infect the lower airways. J. Infect. Dis. 181: Papadopoulos NG, Johnston SL Rhinoviruses as pathogens of the lower respiratory tract. Can. Respir. J. 7: Papadopoulos NG, Psarras S Rhinoviruses in the pathogenesis of asthma. Curr. Allergy Asthma Rep. 3: Rosenthal LA, et al Viral respiratory tract infections and asthma: the course ahead. J. Allergy Clin. Immunol. 125: Schibler M, et al Experimental human rhinovirus and enterovirus interspecies recombination. J. Gen. Virol. 93: Seemungal TA, Harper-Owen R, Bhowmik A, Jeffries DJ, Wedzicha JA Detection of rhinovirus in induced sputum at exacerbation of chronic obstructive pulmonary disease. Eur. Respir. J. 16: Smyth AR, Smyth RL, Tong CY, Hart CA, Heaf DP Effect of respiratory virus infections including rhinovirus on clinical status in cystic fibrosis. Arch. Dis. Child. 73: Tapparel C, et al New molecular detection tools adapted to emerging rhinoviruses and enteroviruses. J. Clin. Microbiol. 47: Tapparel C, et al Rhinovirus genome variation during chronic upper and lower respiratory tract infections. PLoS One 6:e doi: /journal.pone Utokaparch S, et al. The relationship between respiratory viral loads and diagnosis in children presenting to a pediatric hospital emergency department. Pediatr. Infect. Dis. J. 30:e18 e Wat D, et al The role of respiratory viruses in cystic fibrosis. J. Cyst. Fibros. 7: Downloaded from on June 10, 2013 by Bibliotheque Faculte Medecine Geneve 2872 jcm.asm.org Journal of Clinical Microbiology

86 Discussion EXPERIMENTAL INTRA- AND INTERSPECIES RHINOVIRUS RECOMBINATION Recombination, along with a high mutation rate, represents an opportunity for genetic variability and genomic evolution in picornaviruses and other positive-strand RNA viruses. In addition, it allows genetic repair of genomic regions that have degenerated due to extensive mutations accumulation. This molecular process is known to occur frequently among HEV members belonging to the same species and to be a driving force for HEV evolution, as exemplified by the frequent occurrence of cvdpv, resulting from recombination between attenuated oral poliovirus strains and co-circulating non-poliovirus HEV-C strains [58, 59, 61-63]. HEV recombination breakpoints were mainly indentified around the 5 UTR-ORF junction and the P1-2A junction. In contrast, HRV recombination seems to be much more limited [67, 70]. The reason for the difference in recombination rates between these two virus groups belonging to the same genus is not understood. The site of infection and replication could possibly influence recombination opportunities. Also, HRV infections are probably of shorter duration than HEV infection, reducing the chances of recombination. Furthermore, the occurrence of cocirculation of different HEV types might be more frequent than that of different HRV types for unknown reasons, although co-infection with two different HRV types has been identified in respiratory specimens tested in our laboratory (unpublished data). Finally, unknown genomic factors might render HRV recombination less frequent than HEV recombination. 66

87 To further investigate the HRV recombination potential, we took advantage of artificially engineered chimeric HRV constructs and of non replicative recombination to investigate the viability of intraserotypic, intraspecies and interspecies recombinants at the 5 UTR and polyprotein levels. Artificially engineered chimeric HRV genomes The replacement of the HRV-A16 5 UTR by those of HRV-B14, an HRV-C11 strain (HRV- Ca subspecies), an HRV-Cc strain (GenBank accession no. JN087518) and even of HEV-A71 resulted in viable viruses. This suggests that the 5 UTR is largely interchangeable between different HRV species, and even between HRVs and HEVs. An extended number of 5 UTR- ORF chimeras should be tested to assess to what extent this statement is true. Nevertheless, these results are congruent with identified sites of natural HRV recombination close to the 5 UTR-ORF junction [67, 69]. In addition, the 5 UTR region is the most conserved genomic region among HRVs and cannot be used to segregate HRV types into species, which may explain why interspecies recombination is possible in this region. Interestingly, the fittest interspecies recombinant harboured the HEV-A 5 UTR, whose sequence is the most divergent to that of the HRV-A16 5 UTR among the chimeras used in this study. This finding could be explained by the fact that the biologic activity of secondary RNA structures of 5 UTR elements plays a crucial role in HRV translation and replication processes, irrespective of the actual nucleotide sequence. This would imply that the enterovirus 5 UTR per se is particularly efficient in promoting viral translation and replication. See also the discussion section in Schibler et al [162]. Concerning experiments involving chimeric HRVs in which the coding region for the capsid (VP4, V2, VP3 and VP1) and 2A proteins were replaced with corresponding sequences of 67

88 another HRV genome, only the intra-species HRV-A16/HRV-A81 recombinant gave rise to viable and replication competent virus. None of the HRV-A16/HRV-C11, HRV-C11/HRV- A16 and HRV-A16/HRV-B14 interspecies chimeric constructs could be amplified in culture. Furthermore, immunofluorescence performed 7 days post transfection with an antibody detecting the double-stranded replication intermediate was negative, suggesting that the absence of viability of these interspecies chimeric genomes resides upstream of the viral replication process. Of note, as mentioned in the materials and methods section, the nonrecombinant HRV-C genome used for the design of two of our chimeric constructs was unable to replicate upon transfection. Experimental non replicative HRV RNA recombination The non replicative RNA recombination approach was first described by Gmyl et al for poliovirus genomes [73]. However, in these experiments, the authors used deleted poliovirus genomes designed to recombine in the 3 end of the 5 UTR. In this study, the co-transfection partners were not engineered to promote recombination in a particular genomic region. Instead, due to 5 and 3 terminal deletions, recombination could occur at any position in between. As a proof of principle, co-transfection of a 5 deleted HRV-A16 genome with a 3 deleted HRV-A16 genome yielded recombinant viruses in a high proportion of experiments. The two different 5 deleted constructs used provided similar results. As expected, non replicative RNA recombination between HRVs belonging to different types but to the same species was less efficient. Indeed, co-transfection of a 5 UTR deleted HRV- A16 genome and a 3 deleted HRV-A39 genome resulted in viable recombination in two out of eight wells. Repeating these experiments allowed us to recover two additional intraspecies recombinants. Mapping of the recombination sites revealed breakpoints in VP2, VP1 and 3D. 68

89 The identification of recombination sites in the capsid genes was rather surprising, as this has never been described in natural HRV intraspecies recombination. This may be linked to a different recombination mechanism occurring in vivo, namely replicative recombination, or to a specific environmental pressure preventing the emergence of HRVs having recombined in the region coding for capsid proteins. Finally, the parental genomes are not defective in vivo and may outcompete recombinants at the capsid level. Nevertheless, these results suggest that chimeric capsids originating from different serotypes within a same species are potentially functional. As for the artificially engineered chimeras, no inter-species recombinant could be recovered by non replicative recombination. However, we observed clusters of a few IFpositive cells after HRV-C11 del3 and HRV-A16 del5 UTR (1-434) suggesting that viable non replicative RNA recombination did occur, but the resulting virus was not fit enough to be successfully passaged and sequenced. Of note, though these data are in agreement with data obtained with chimeric genomes, solid conclusions are difficult to draw since the full-length HRV-C11 genome was unable to replicate and was initially used in an attempt to understand why HRV-Cs could not be propagated in cell cultures. Altogether the artificial and non replicative recombinants generated in this study suggest that although 5 UTR interspecies recombinant are viable, the recombination potential at the polyprotein level is much more limited and possible only within a serotype or to a lesser extent within a given HRV species. These global findings are in agreement with observations derived from circulating HRV recombinants [67]. Interspecies recombination in the polyprotein region did not result in fully infective virions regardless of the experimental approach used. There are at least two constraints that may limit interspecies recombination. First, as polyprotein cleavage sites differ among the three HRV species, a likely hypothesis explaining the absence of viable interspecies HRV recombinants 69

90 at the polyprotein region level is that the 2A and 3C proteases of a given HRV species can only fully process polyproteins translated from genomes of the same species. In experiments performed in this study 2A protease cleavage incompatibilities were avoided by exchanging the 2A gene along with the capsid genes originating from the same HRV genome. Concerning 3C protease cleavage sites, we identified one in the HRV-B14 capsid region that differs from the corresponding site in the HRV A16 polyprotein, between the VP3 and VP1 regions (Table 3). Hence, this site in the HRV-B14 capsid region might not be appropriately cleaved by the HRV-A16 3C protease, possibly explaining why the HRV-A16/HRV-B14 P12A recombinant did not yield viable virus. Second, the fact that cre elements, essential for HRV replication, are located in different ORF regions according to the different HRV species constitutes another obstacle susceptible to limit interspecies HRV recombination in vivo. A substantial proportion of theoretical interspecies HRV recombination combinations could therefore result in the absence of a functional cre element, rendering the recombinant HRV genome unable to replicate. On the other hand, an additional cre may arise from interspecies HRV recombination, which might interfere with the replication process. Our engineered P1-2A HRV recombinants all displayed a single and functional cre structure, which theoretically allows the replication process to occur properly. However, potential replication disturbances related to the displacement of this element in interspecies P1-2A chimeras cannot be formally excluded. The absence of interspecies HRV recombination in these two systems is somewhat in contradiction with the HRV interspecies recombinants documented in the literature [69]. This may rely on the fact that all these documented recombination events are ancient and might have occurred before the individual speciation of the parental HRV genomes. 70

91 HRV RNA QUANTIFICATION BY REAL-TIME RT-PCR IN RESPIRATORY SPECIMENS The second part of this thesis concerns the attempt and feasibility of HRV RNA quantification by real-time RT-PCR in respiratory specimens, which is discussed in [163]. The Panenterhino/Ge/08 real-time RT-PCR assay proved to be highly reproducible and therefore allowed reliable estimation of HRV RNA load in a respiratory specimen. The experimentally estimated error margin related to the quantification of RNAs derived from genetically diverse HRVs was surprisingly low, below 10% of the quantification value. However, due to the variability linked to the use of different assays and RNA controls, universal absolute RNA quantification in clinical specimens is currently not feasible. The accurate quantification of a commercial HRV RNA standard could circumvent this limitation. This problem could also be avoided by the advent of digital PCR. Indeed, in this recent technology, the sample is partitioned into numerous micro wells so that each of them contains either one or no nucleic acid molecule. Therefore, after PCR amplification, nucleic acids can be quantified by counting the positive reactions, allowing absolute nucleic acid quantification without the requirement of an RNA standard. Nevertheless, in our opinion, the major limitation regarding HRV RNA quantification in respiratory specimens resides in the variability of sampling procedures and the time point at which the sampling is performed. Indeed, such quantification is likely to vary significantly depending on how deep the swab penetrates the nose and how much force is applied to the nasal or pharyngeal mucosa. It also seems difficult to compare RNA quantification results obtained from different respiratory tract sites (nasal swabs, tracheal aspirates, BAL, etc.). Furthermore, the quantity of virus varies significantly during the course of an acute HRV infection. 71

92 CONCLUSIONS AND PROSPECTS The work presented in this thesis focused on two diverse aspects of rhinoviruses, which represent a frequent human pathogen. The first part addressed a basic science theme of virology, namely recombination. Experimental intra- and interspecies HRV recombination using artificially engineered chimeras and non replicative recombination was studied. The small number of recombinants comprised in our results and the usage of a growth deficient parental HRV-C render it difficult to draw general conclusions regarding HRV recombination patterns. Nevertheless, the two different and somehow complementary experimental approaches used in this study provided some valuable information, which may be summarized as follows. First, the 5 UTR of HRVs, and even HEVs, seem to be largely interchangeable. Second, recombination in the coding region generated viable intraspecies but not interspecies recombinants. Third, non replicative intraspecies RNA recombination occurs in cell culture upon co-transfection of two complementary deleted HRV genomes and this method could be used to map theoretical HRV recombination sites. The application of these approaches to an extended number of combinations could expand our understanding of HRV genomic organization and perhaps of genetic determinants of specific phenotypes. HRV chimeras in which the P2 and P3 regions, coding for non structural proteins, originate from two different HRV genomes could be engineered and tested. Intraspecies chimeras representative of each of the three HRV species may be studied in order to obtain a comprehensive understanding of the experimental HRV recombination potential. Interspecies recombinants in the same genomic regions could also be designed to test the HRV interspecies genetic barrier. Additional non replicative HRV RNA recombination experiments including co-transfection of partially deleted but complementary HRV RNA pairs of each 72

93 HRV species would certainly add supplementary knowledge. Of course, experiments involving HRV-C recombinants will require appropriate culture systems, as members of this species are not cultivable in classic cell lines. The second part of this thesis concerned a more clinical aspect of rhinoviruses, namely the feasibility of HRV RNA quantification in respiratory specimens using real-time RT-PCR. Although the Panenterhino/Ge/08 real-time RT-PCR assay was accurate for HRV RNA quantification, the results obtained should be interpreted carefully, according to the issues discussed above. However, this tool might be of use in specific clinical situations, provided that sampling methods are rigorously standardized. For instance, subactute or chronic HRV infections in immunosupressed patients could be monitored, and correlations between viral load and symptom severity could be assessed with this assay. Clinical studies addressing these issues are currently underway. In the future, measuring HRV RNA loads could assess the efficacy of anti HRV drugs. 73

94 References 1. Reuter, G., A. Boros, and P. Pankovics, Kobuviruses - a comprehensive review. Rev Med Virol. 21(1): p Rhinoviruses: a numbering system. Nature, (5078): p Hamparian, V.V., et al., A collaborative report: rhinoviruses--extension of the numbering system from 89 to 100. Virology, (1): p Savolainen, C., et al., Genetic clustering of all 102 human rhinovirus prototype strains: serotype 87 is close to human enterovirus 70. J Gen Virol, (Pt 2): p Wisdom, A., et al., Screening respiratory samples for detection of human rhinoviruses (HRVs) and enteroviruses: comprehensive VP4-VP2 typing reveals high incidence and genetic diversity of HRV species C. J Clin Microbiol, (12): p Simmonds, P., et al., Proposals for the classification of human rhinovirus species C into genotypically assigned types. J Gen Virol. 91(Pt 10): p Savolainen-Kopra, C., et al., 5' noncoding region alone does not unequivocally determine genetic type of human rhinovirus strains. J Clin Microbiol, (4): p Tapparel, C., et al., Rhinovirus genome variation during chronic upper and lower respiratory tract infections. PLoS One. 6(6): p. e Sherry, B., et al., Use of monoclonal antibodies to identify four neutralization immunogens on a common cold picornavirus, human rhinovirus 14. J Virol, (1): p

95 10. Verdaguer, N., D. Blaas, and I. Fita, Structure of human rhinovirus serotype 2 (HRV2). J Mol Biol, (5): p McLean, G.R., et al., Rhinovirus infections and immunisation induce cross-serotype reactive antibodies to VP1. Antiviral Res. 95(3): p Arnold, E. and M.G. Rossmann, Analysis of the structure of a common cold virus, human rhinovirus 14, refined at a resolution of 3.0 A. J Mol Biol, (4): p Olson, N.H., et al., Structure of a human rhinovirus complexed with its receptor molecule. Proc Natl Acad Sci U S A, (2): p Colonno, R.J., et al., Evidence for the direct involvement of the rhinovirus canyon in receptor binding. Proc Natl Acad Sci U S A, (15): p Hadfield, A.T., G.D. Diana, and M.G. Rossmann, Analysis of three structurally related antiviral compounds in complex with human rhinovirus 16. Proc Natl Acad Sci U S A, (26): p Rossmann, M.G., The canyon hypothesis. Hiding the host cell receptor attachment site on a viral surface from immune surveillance. J Biol Chem, (25): p Cordey, S., et al., Rhinovirus : diversité clinique et génomique. Virologie, (5): p Cordey, S., et al., The cis-acting replication elements define human enterovirus and rhinovirus species. RNA, (8): p Rohll, J.B., et al., The 5'-untranslated regions of picornavirus RNAs contain independent functional domains essential for RNA replication and translation. J Virol, (7): p

96 20. Gamarnik, A.V. and R. Andino, Switch from translation to RNA replication in a positive-stranded RNA virus. Genes Dev, (15): p Perera, R., et al., Cellular protein modification by poliovirus: the two faces of poly(rc)-binding protein. J Virol, (17): p Borman, A.M., F.G. Deliat, and K.M. Kean, Sequences within the poliovirus internal ribosome entry segment control viral RNA synthesis. EMBO J, (13): p Brown, D.M., et al., An authentic 3' noncoding region is necessary for efficient poliovirus replication. J Virol, (18): p Steil, B.P. and D.J. Barton, Cis-active RNA elements (CREs) and picornavirus RNA replication. Virus Res, (2): p Lin, J.Y., et al., Viral and host proteins involved in picornavirus life cycle. J Biomed Sci, : p Vlasak, M., et al., The minor receptor group of human rhinovirus (HRV) includes HRV23 and HRV25, but the presence of a lysine in the VP1 HI loop is not sufficient for receptor binding. J Virol, (12): p Bochkov, Y.A. and J.E. Gern, Clinical and molecular features of human rhinovirus C. Microbes Infect. 14(6): p Hao, W., et al., Infection and propagation of human rhinovirus C in human airway epithelial cells. J Virol. 86(24): p McErlean, P., et al., Distinguishing molecular features and clinical characteristics of a putative new rhinovirus species, human rhinovirus C (HRV C). PLoS One, (4): p. e Fuchs, R. and D. Blaas, Uncoating of human rhinoviruses. Rev Med Virol. 20(5): p

97 31. Rozovics, J.M., R. Virgen-Slane, and B.L. Semler, Engineered picornavirus VPg-RNA substrates: analysis of a tyrosyl-rna phosphodiesterase activity. PLoS One. 6(3): p. e Niepmann, M., Internal translation initiation of picornaviruses and hepatitis C virus. Biochim Biophys Acta, (9-10): p Etchison, D. and S. Fout, Human rhinovirus 14 infection of HeLa cells results in the proteolytic cleavage of the p220 cap-binding complex subunit and inactivates globin mrna translation in vitro. J Virol, (2): p Pacheco, A. and E. Martinez-Salas, Insights into the biology of IRES elements through riboproteomic approaches. J Biomed Biotechnol. 2010: p Martinez-Salas, E. and M. Ryan, Translation and Protein Processing, in The Picornaviruses, E. Ehrenfeld, E. Domingo, and R. Roos, Editors. 2010, ASM Press: Washington DC. p Arnold, E., et al., Implications of the picornavirus capsid structure for polyprotein processing. Proc Natl Acad Sci U S A, (1): p Bedard, K.M. and B.L. Semler, Regulation of picornavirus gene expression. Microbes Infect, (7): p Herold, J. and R. Andino, Poliovirus RNA replication requires genome circularization through a protein-protein bridge. Mol Cell, (3): p Todd, S., et al., Replication-competent picornaviruses with complete genomic RNA 3' noncoding region deletions. J Virol, (11): p Silvestri, L.S., et al., Relationship between poliovirus negative-strand RNA synthesis and the length of the 3' poly(a) tail. Virology, (2): p Ogram, S.A. and J.B. Flanegan, Non-template functions of viral RNA in picornavirus replication. Curr Opin Virol. 1(5): p

98 42. Novak, J.E. and K. Kirkegaard, Improved method for detecting poliovirus negative strands used to demonstrate specificity of positive-strand encapsidation and the ratio of positive to negative strands in infected cells. J Virol, (6): p Bishop, J.M., D.F. Summers, and L. Levintow, Characterization of ribonucleaseresistant RNA from poliovirus-infected HeLa cells. Proc Natl Acad Sci U S A, (4): p Goodfellow, I.G., et al., The poliovirus 2C cis-acting replication element-mediated uridylylation of VPg is not required for synthesis of negative-sense genomes. J Gen Virol, (Pt 9): p Morasco, B.J., et al., Poliovirus cre(2c)-dependent synthesis of VPgpUpU is required for positive- but not negative-strand RNA synthesis. J Virol, (9): p McKnight, K.L. and S.M. Lemon, The rhinovirus type 14 genome contains an internally located RNA structure that is required for viral replication. RNA, (12): p Yang, Y., et al., Sequence requirements for viral RNA replication and VPg uridylylation directed by the internal cis-acting replication element (cre) of human rhinovirus type 14. J Virol, (15): p Salonen, A., T. Ahola, and L. Kaariainen, Viral RNA replication in association with cellular membranes. Curr Top Microbiol Immunol, : p Jackson, W.T., et al., Subversion of cellular autophagosomal machinery by RNA viruses. PLoS Biol, (5): p. e Teterina, N.L., et al., Poliovirus 2C protein determinants of membrane binding and rearrangements in mammalian cells. J Virol, (12): p Quiner, C.A. and W.T. Jackson, Fragmentation of the Golgi apparatus provides replication membranes for human rhinovirus 1A. Virology. 407(2): p

99 52. Lee, W.M. and W. Wang, Human rhinovirus type 16: mutant V1210A requires capsidbinding drug for assembly of pentamers to form virions during morphogenesis. J Virol, (11): p Compton, S.R., B. Nelsen, and K. Kirkegaard, Temperature-sensitive poliovirus mutant fails to cleave VP0 and accumulates provirions. J Virol, (9): p Racaniello, V.R., Picornaviridae: The Viruses and Their Replication, in Fields Virology, D. Knipe and P. Howley, Editors. 2006, Lippincott Williams & Wilkins: Phildelphia. p Winther, B., et al., Light and scanning electron microscopy of nasal biopsy material from patients with naturally acquired common colds. Acta Otolaryngol, (3-4): p Domingo, E. and J.J. Holland, RNA virus mutations and fitness for survival. Annu Rev Microbiol, : p Lauring, A.S. and R. Andino, Quasispecies theory and the behavior of RNA viruses. PLoS Pathog. 6(7): p. e Lukashev, A.N., Role of recombination in evolution of enteroviruses. Rev Med Virol, (3): p Santti, J., et al., Evidence of recombination among enteroviruses. J Virol, (10): p Simmonds, P. and J. Welch, Frequency and dynamics of recombination within different species of human enteroviruses. J Virol, (1): p Joffret, M.L., et al., Common and diverse features of cocirculating type 2 and 3 recombinant vaccine-derived polioviruses isolated from patients with poliomyelitis and healthy children. J Infect Dis, (9): p

100 62. Jegouic, S., et al., Recombination between polioviruses and co-circulating Coxsackie A viruses: role in the emergence of pathogenic vaccine-derived polioviruses. PLoS Pathog, (5): p. e Rousset, D., et al., Recombinant vaccine-derived poliovirus in Madagascar. Emerg Infect Dis, (7): p Yozwiak, N.L., et al., Human enterovirus 109: a novel interspecies recombinant enterovirus isolated from a case of acute pediatric respiratory illness in Nicaragua. J Virol. 84(18): p Bolanaki, E., et al., Partial 3D gene sequences of Coxsackie viruses reveal interspecies exchanges. Virus Genes, (2): p Smura, T., et al., Enterovirus surveillance reveals proposed new serotypes and provides new insight into enterovirus 5'-untranslated region evolution. J Gen Virol, (Pt 9): p Tapparel, C., et al., New respiratory enterovirus and recombinant rhinoviruses among circulating picornaviruses. Emerg Infect Dis, (5): p Huang, T., et al., Evidence of recombination and genetic diversity in human rhinoviruses in children with acute respiratory infection. PLoS One, (7): p. e McIntyre, C.L., et al., Analysis of genetic diversity and sites of recombination in human rhinovirus species C. J Virol. 84(19): p Palmenberg, A.C., et al., Sequencing and analyses of all known human rhinovirus genomes reveal structure and evolution. Science, (5923): p Tapparel, C., et al., New complete genome sequences of human rhinoviruses shed light on their phylogeny and genomic features. BMC Genomics, : p

101 72. Lukashev, A.N., Recombination among picornaviruses. Rev Med Virol. 20(5): p Gmyl, A.P., et al., Nonreplicative RNA recombination in poliovirus. J Virol, (11): p Simon-Loriere, E. and E.C. Holmes, Why do RNA viruses recombine? Nat Rev Microbiol. 9(8): p Agol, V.I., Picornavirus Genetics: an Overview, in Molecular Biology of Picornaviruses, B.L. Semler and E. Wimmer, Editors. 2002, ASM Press: Washington DC. p Oberste, M.S., et al., Enterovirus 68 is associated with respiratory illness and shares biological features with both the enteroviruses and the rhinoviruses. J Gen Virol, (Pt 9): p Papadopoulos, N.G., et al., Rhinoviruses replicate effectively at lower airway temperatures. J Med Virol, (1): p Couch, R.B., Rhinoviruses, in Fields Virology, D. Knipe and P. Howley, Editors. 2001, Lippincott Williams & Wilkins: Philadelphia. p Arruda, E., et al., Comparative susceptibilities of human embryonic fibroblasts and HeLa cells for isolation of human rhinoviruses. J Clin Microbiol, (5): p Lewis, F.A. and M.L. Kennett, Comparison of rhinovirus-sensitive HeLa cells and human embryo fibroblasts for isolation of rhinoviruses from patients with respiratory disease. J Clin Microbiol, (5): p Terletskaia-Ladwig, E., et al., A convenient rapid culture assay for the detection of enteroviruses in clinical samples: comparison with conventional cell culture and RT- PCR. J Med Microbiol, (Pt 8): p

102 82. Tsao, K.C., et al., Epidemiologic features and virus isolation of enteroviruses in Northern Taiwan during J Virol Methods. 165(2): p Hendley, J.O., R.P. Wenzel, and J.M. Gwaltney, Jr., Transmission of rhinovirus colds by self-inoculation. N Engl J Med, (26): p Dick, E.C., et al., Aerosol transmission of rhinovirus colds. J Infect Dis, (3): p Winther, B., Rhinovirus infections in the upper airway. Proc Am Thorac Soc. 8(1): p Arruda, E., et al., Localization of human rhinovirus replication in the upper respiratory tract by in situ hybridization. J Infect Dis, (5): p Kennedy, J.L., et al., Pathogenesis of rhinovirus infection. Curr Opin Virol. 2(3): p Feng, Q., et al., MDA5 detects the double-stranded RNA replicative form in picornavirus-infected cells. Cell Rep. 2(5): p Kato, H., et al., Length-dependent recognition of double-stranded ribonucleic acids by retinoic acid-inducible gene-i and melanoma differentiation-associated gene 5. J Exp Med, (7): p Pichlmair, A. and C. Reis e Sousa, Innate recognition of viruses. Immunity, (3): p Takeuchi, O. and S. Akira, MDA5/RIG-I and virus recognition. Curr Opin Immunol, (1): p Darnell, J.E., Jr., I.M. Kerr, and G.R. Stark, Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science, (5164): p

103 93. Der, S.D., et al., Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc Natl Acad Sci U S A, (26): p Taniguchi, T. and A. Takaoka, A weak signal for strong responses: interferonalpha/beta revisited. Nat Rev Mol Cell Biol, (5): p Barral, P.M., et al., MDA-5 is cleaved in poliovirus-infected cells. J Virol, (8): p Drahos, J. and V.R. Racaniello, Cleavage of IPS-1 in cells infected with human rhinovirus. J Virol, (22): p Morrison, J.M. and V.R. Racaniello, Proteinase 2Apro is essential for enterovirus replica tion in type I interferon-treated cells. J Virol, (9): p Kelly, J.T. and W.W. Busse, Host immune responses to rhinovirus: mechanisms in asthma. J Allergy Clin Immunol, (4): p ; quiz Gern, J.E., et al., Rhinovirus-specific T cells recognize both shared and serotyperestricted viral epitopes. J Infect Dis, (5): p Barclay, W.S., et al., The time course of the humoral immune response to rhinovirus infection. Epidemiol Infect, (3): p Cordey, S., et al., Rhinovirus genome evolution during experimental human infection. PLoS One. 5(5): p. e Alper, C.M., et al., Prechallenge antibodies moderate disease expression in adults experimentally exposed to rhinovirus strain hanks. Clin Infect Dis, (1): p Alper, C.M., et al., Prechallenge antibodies: moderators of infection rate, signs, and symptoms in adults experimentally challenged with rhinovirus type 39. Laryngoscope, (10): p

104 104. Kennedy, J.L., et al., Pathogenesis of rhinovirus infection. Curr Opin Virol, (3): p Jackson, D.J. and S.L. Johnston, The role of viruses in acute exacerbations of asthma. J Allergy Clin Immunol. 125(6): p ; quiz Heikkinen, T. and A. Jarvinen, The common cold. Lancet, (9351): p Hendley, J.O., The host response, not the virus, causes the symptoms of the common cold. Clin Infect Dis, (4): p Iwane, M.K., et al., Human rhinovirus species associated with hospitalizations for acute respiratory illness in young US children. J Infect Dis. 204(11): p Jin, Y., et al., Prevalence and clinical characterization of a newly identified human rhinovirus C species in children with acute respiratory tract infections. J Clin Microbiol, (9): p Kiang, D., et al., Molecular characterization of a variant rhinovirus from an outbreak associated with uncommonly high mortality. J Clin Virol, (3): p Lau, S.K., et al., Clinical features and complete genome characterization of a distinct human rhinovirus (HRV) genetic cluster, probably representing a previously undetected HRV species, HRV-C, associated with acute respiratory illness in children. J Clin Microbiol, (11): p Lee, W.M., et al., Human rhinovirus species and season of infection determine illness severity. Am J Respir Crit Care Med. 186(9): p Gwaltney, J.M., Jr., et al., Computed tomographic study of the common cold. N Engl J Med, (1): p Chantzi, F.M., et al., Human rhinoviruses in otitis media with effusion. Pediatr Allergy Immunol, (7): p

105 115. Pitkaranta, A., et al., Polymerase chain reaction-based detection of rhinovirus, respiratory syncytial virus, and coronavirus in otitis media with effusion. J Pediatr, (3): p Stott, E.J. and G.F. Heath, Factors affecting the growth of Rhinovirus 2 in suspension cultures of L132 cells. J Gen Virol, (1): p Cate, T.R., et al., Production of Tracheobronchitis in Volunteers with Rhinovirus in a Small-Particle Aerosol. Am J Epidemiol, : p Craighead, J.E., M. Meier, and M.H. Cooley, Pulmonary infection due to rhinovirus type 13. N Engl J Med, (25): p Gern, J.E., et al., Detection of rhinovirus RNA in lower airway cells during experimentally induced infection. Am J Respir Crit Care Med, (3): p Papadopoulos, N.G. and S.L. Johnston, Rhinoviruses as pathogens of the lower respiratory tract. Can Respir J, (5): p Kaiser, L., et al., Chronic rhinoviral infection in lung transplant recipients. Am J Respir Crit Care Med, (12): p Malmstrom, K., et al., Human rhinovirus in bronchial epithelium of infants with recurrent respiratory symptoms. J Allergy Clin Immunol, (3): p Peltola, V., et al., Rhinovirus infections in children: a retrospective and prospective hospital-based study. J Med Virol, (10): p Seemungal, T.A., et al., Detection of rhinovirus in induced sputum at exacerbation of chronic obstructive pulmonary disease. Eur Respir J, (4): p Papadopoulos, N.G. and S. Psarras, Rhinoviruses in the pathogenesis of asthma. Curr Allergy Asthma Rep, (2): p

106 126. Rosenthal, L.A., et al., Viral respiratory tract infections and asthma: the course ahead. J Allergy Clin Immunol. 125(6): p de Almeida, M.B., et al., Rhinovirus C and respiratory exacerbations in children with cystic fibrosis. Emerg Infect Dis. 16(6): p Smyth, A.R., et al., Effect of respiratory virus infections including rhinovirus on clinical status in cystic fibrosis. Arch Dis Child, (2): p Wat, D., et al., The role of respiratory viruses in cystic fibrosis. J Cyst Fibros, (4): p Wark, P.A., et al., Asthmatic bronchial epithelial cells have a deficient innate immune response to infection with rhinovirus. J Exp Med, (6): p Graat, J.M., et al., A prospective, community-based study on virologic assessment among elderly people with and without symptoms of acute respiratory infection. J Clin Epidemiol, (12): p Peltola, V., et al., Rhinovirus transmission within families with children: incidence of symptomatic and asymptomatic infections. J Infect Dis, (3): p Winther, B., F.G. Hayden, and J.O. Hendley, Picornavirus infections in children diagnosed by RT-PCR during longitudinal surveillance with weekly sampling: Association with symptomatic illness and effect of season. J Med Virol, (5): p Johnston, S.L., et al., Use of polymerase chain reaction for diagnosis of picornavirus infection in subjects with and without respiratory symptoms. J Clin Microbiol, (1): p Mackay, I.M., Human rhinoviruses: the cold wars resume. J Clin Virol, (4): p

107 136. Arden, K.E. and I.M. Mackay, Newly identified human rhinoviruses: molecular methods heat up the cold viruses. Rev Med Virol. 20(3): p Atmar, R. and S. Greenberg, Lennette's Laboratory Diagnosis of Viral Infections. Fourth ed Gama, R.E., et al., Amplification of rhinovirus specific nucleic acids from clinical samples using the polymerase chain reaction. J Med Virol, (2): p Gama, R.E., et al., Polymerase chain reaction amplification of rhinovirus nucleic acids from clinical material. Nucleic Acids Res, (19): p Hyypia, T., P. Auvinen, and M. Maaronen, Polymerase chain reaction for human picornaviruses. J Gen Virol, ( Pt 12): p Dagher, H., et al., Rhinovirus detection: comparison of real-time and conventional PCR. J Virol Methods, (2): p Kares, S., et al., Real-time PCR for rapid diagnosis of entero- and rhinovirus infections using LightCycler. J Clin Virol, (2): p Tapparel, C., et al., New molecular detection tools adapted to emerging rhinoviruses and enteroviruses. J Clin Microbiol, (6): p Gambarino, S., et al., Development of a RT real-time PCR for the detection and quantification of human rhinoviruses. Mol Biotechnol, (3): p Bardin, P.G., et al., Detection of rhinovirus infection of the nasal mucosa by oligonucleotide in situ hybridization. Am J Respir Cell Mol Biol, (2): p Pitkaranta, A., et al., Rhinovirus RNA in the maxillary sinus epithelium of adult patients with acute sinusitis. Clin Infect Dis, (6): p

108 147. Loens, K., et al., Improved detection of rhinoviruses by nucleic acid sequence-based amplification after nucleotide sequence determination of the 5' noncoding regions of additional rhinovirus strains. J Clin Microbiol, (5): p Samuelson, A., et al., Development and application of a new method for amplification and detection of human rhinovirus RNA. J Virol Methods, (2): p Kistler, A., et al., Pan-viral screening of respiratory tract infections in adults with and without asthma reveals unexpected human coronavirus and human rhinovirus diversity. J Infect Dis, (6): p Ecker, D.J., et al., Ibis T5000: a universal biosensor approach for microbiology. Nat Rev Microbiol, (7): p Scott, G.M., et al., Purified interferon as protection against rhinovirus infection. Br Med J (Clin Res Ed), (6332): p Tyrrell, D.A., Trials of interferon in respiratory infections of man. Tex Rep Biol Med, : p Turner, R.B., et al., Efficacy of tremacamra, a soluble intercellular adhesion molecule 1, for experimental rhinovirus infection: a randomized clinical trial. JAMA, (19): p Hayden, F.G., et al., Efficacy and safety of oral pleconaril for treatment of colds due to picornaviruses in adults: results of 2 double-blind, randomized, placebo-controlled trials. Clin Infect Dis, (12): p Kaiser, L., C.E. Crump, and F.G. Hayden, In vitro activity of pleconaril and AG7088 against selected serotypes and clinical isolates of human rhinoviruses. Antiviral Res, (3): p Pevear, D.C., et al., Activity of pleconaril against enteroviruses. Antimicrob Agents Chemother, (9): p

109 157. Rotbart, H.A. and A.D. Webster, Treatment of potentially life-threatening enterovirus infections with pleconaril. Clin Infect Dis, (2): p Hayden, F.G., et al., Phase II, randomized, double-blind, placebo-controlled studies of ruprintrivir nasal spray 2-percent suspension for prevention and treatment of experimentally induced rhinovirus colds in healthy volunteers. Antimicrob Agents Chemother, (12): p Wanga, Q.M. and S.H. Chen, Human rhinovirus 3C protease as a potential target for the development of antiviral agents. Curr Protein Pept Sci, (1): p Thibaut, H.J., A.M. De Palma, and J. Neyts, Combating enterovirus replication: stateof-the-art on antiviral research. Biochem Pharmacol. 83(2): p Edlmayr, J., et al., Antibodies induced with recombinant VP1 from human rhinovirus exhibit cross-neutralisation. Eur Respir J. 37(1): p Schibler, M., et al., Experimental human rhinovirus and enterovirus interspecies recombination. J Gen Virol. 93(Pt 1): p Schibler, M., et al., A critical analysis of rhinovirus RNA load quantification by realtime RT-PCR. J Clin Microbiol Tapparel, C., et al., Pneumonia and pericarditis in a child with HRV-C infection: a case report. J Clin Virol, (2): p Jurgeit, A., et al., An RNA replication-center assay for high content image-based quantifications of human rhinovirus and coxsackievirus infections. Virol J. 7: p

110 Appendix ARTICLE 3 Identification of Site-Specific Adaptations Conferring Increased Neural Cell Tropism during Human Enterovirus 71 Infection. Cordey S, Petty TJ, Schibler M, Martinez Y, Gerlach D, van Belle S, Turin L, Zdobnov E, Kaiser L, Tapparel C. PLoS Pathog Jul;8(7):e Epub 2012 Jul

111 Identification of Site-Specific Adaptations Conferring Increased Neural Cell Tropism during Human Enterovirus 71 Infection Samuel Cordey 1,2. *, Tom J. Petty 3,4., Manuel Schibler 1,2, Yannick Martinez 5, Daniel Gerlach 6, Sandra van Belle 1,2, Lara Turin 1,2, Evgeny Zdobnov 3,4,7, Laurent Kaiser 1,2", Caroline Tapparel 1,2" 1 Laboratory of Virology, Division of Infectious Diseases and Division of Laboratory Medicine, University Hospitals of Geneva, Geneva, Switzerland, 2 Department of Medicine, University of Geneva Medical School, Geneva, Switzerland, 3 Department of Genetic Medicine and Development, University of Geneva Medical School, Geneva, Switzerland, 4 Swiss Institute of Bioinformatics, Geneva, Switzerland, 5 Department of Pathology and Immunology, Faculty of Medicine, University of Geneva, Geneva, Switzerland, 6 Research Institute of Molecular Pathology (IMP), Vienna, Austria, 7 Imperial College London, South Kensington Campus, London, United Kingdom Abstract Enterovirus 71 (EV71) is one of the most virulent enteroviruses, but the specific molecular features that enhance its ability to disseminate in humans remain unknown. We analyzed the genomic features of EV71 in an immunocompromised host with disseminated disease according to the different sites of infection. Comparison of five full-length genomes sequenced directly from respiratory, gastrointestinal, nervous system, and blood specimens revealed three nucleotide changes that occurred within a five-day period: a non-conservative amino acid change in VP1 located within the BC loop (L97R), a region considered as an immunogenic site and possibly important in poliovirus host adaptation; a conservative amino acid substitution in protein 2B (A38V); and a silent mutation in protein 3D (L175). Infectious clones were constructed using both BrCr (lineage A) and the clinical strain (lineage C) backgrounds containing either one or both non-synonymous mutations. In vitro cell tropism and competition assays revealed that the VP1 97 Leu to Arg substitution within the BC loop conferred a replicative advantage in SH-SY5Y cells of neuroblastoma origin. Interestingly, this mutation was frequently associated in vitro with a second non-conservative mutation (E167G or E167A) in the VP1 EF loop in neuroblastoma cells. Comparative models of these EV71 VP1 variants were built to determine how the substitutions might affect VP1 structure and/or interactions with host cells and suggest that, while no significant structural changes were observed, the substitutions may alter interactions with host cell receptors. Taken together, our results show that the VP1 BC loop region of EV71 plays a critical role in cell tropism independent of EV71 lineage and, thus, may have contributed to dissemination and neurotropism in the immunocompromised patient. Citation: Cordey S, Petty TJ, Schibler M, Martinez Y, Gerlach D, et al. (2012) Identification of Site-Specific Adaptations Conferring Increased Neural Cell Tropism during Human Enterovirus 71 Infection. PLoS Pathog 8(7): e doi: /journal.ppat Editor: Bert L. Semler, University of California, Irvine, United States of America Received January 24, 2012; Accepted June 16, 2012; Published July 26, 2012 Copyright: ß 2012 Cordey et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This study was supported by the Swiss National Science Foundation (Grant No to CT and 32003B to LK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * samuel.cordey@hcuge.ch. These authors contributed equally to this work. " These authors also contributed equally to this work. Introduction In humans, enteroviruses target a variety of different organs causing gastrointestinal, respiratory, myocardial, and central nervous system (CNS) diseases [1,2]. The ability of enteroviruses other than poliovirus to cause neurological complications is restricted to a limited number of serotypes that include enterovirus 71 (EV71) [3,4]. EV71 is of particular interest since it can cause major hand-foot-and-mouth disease outbreaks, such as those recently reported across the Asia-Pacific countries [5 8]. Nevertheless, EV71 dissemination to the CNS remains a rare event, as demonstrated by the relatively small proportion of meningoencephalitis among millions of hand-foot-and-mouth disease cases [9 12]. For poliovirus, CNS invasion is thought to occur either through disruption of the blood-brain barrier or via retrograde axonal transport [8]. For EV71, experimental studies in mouse models using adapted strains suggest that the virus has the propensity to invade the CNS through retrograde axonal transport and that hematogenous transport might represent only a minor route of transmission [13 15]. However, the observations in mouse models do not necessarily reflect how CNS invasion occurs during human infections. Neutrotropic enteroviruses need to escape the host defences to reach the CNS. The absence of pre-existing protective immunity, together with a relatively deficient innate immunity, is considered as the first step toward high blood viremia that will then lead to a secondary invasion of the CNS [16]. This explains why young children present more severe diseases. An inefficient immune response could also be the result of a high inoculum size, leading to an overwhelming replication and viremia. However, neurotropism is a multistep event that requires the virus not only to sustain high PLoS Pathogens 1 July 2012 Volume 8 Issue 7 e

112 Disseminated Human Enterovirus 71 Infection Author Summary Human enterovirus-71 (EV71) has been the cause of major hand-foot-and-mouth disease outbreaks, particularly in the Asia-Pacific region. EV71 infection can also disseminate to the central nervous system and result in meningoencephalitis. Despite intensive epidemiological screening, as well as experimentation in animal models, viral factors contributing to neurotropism remain ill-defined. We describe here the analysis of the full-length genomes of EV71 from different infection sites in an immunocompromised host with disseminated disease. Our data highlight a critical amino acid change within the EV71 VP1 protein that could potentially lead to dissemination and neurotropism during natural infections. This hypothesis was confirmed in vitro through reverse genetic experiments in different EV71 lineages and by in silico modelling. To our knowledge, this study provides the first genome-wide analysis of EV71 evolution and dissemination within a single human host over the course of an infection, and highlights how the emergence of mutations at critical regions of the viral genome can potentially lead to new phenotypes and neurovirulence. replication levels, but also to locate a permissive cell type within the CNS. Viral factors contributing to neurotropism have been intensively studied in vitro and in animal models in vivo using poliovirus or non-polio EVs [15 23], but still remain ill-defined. Until now, to the best of our knowledge, EV71 virulence factors and adaptation have not been studied directly from clinical samples during natural human infections and it remains unknown whether secondary seeding from the primary site is only a fortuitous event or if it is associated with specific viral genomic adaptation within the human host. In this study, we analyzed the genomes of EV71 from different sites of infection in an immunocompromised host with disseminated disease. This provided a unique opportunity to investigate any potential intra-host adaptation following natural human infection and to assess whether enterovirus needs to harbor specific genomic features in order to sustain dissemination. After sequence analysis of the collected specimens, amino acid changes observed in the viral proteins VP1 and 2B and possibly associated with neurotropism were further studied both in vitro using a series of different constructs and in silico using comparative models of EV71 VP1. Results Case Description A 38-year-old man with chronic lymphocytic leukemia and recently treated with four courses of chemotherapy, including rituximab, was hospitalized with fever and respiratory symptoms. Five days before admission, he developed fever (39uC), odynophagia, chills, dyspnoea with wheezing, cough and sputum. The total immoglobulin G level in blood was low at 2.74 g/l (normal range, g/l), as were the IgM (0.1 g/l; normal range, ) and IgA (0.17 g/l; normal range, ) levels. Despite intravenous wide-spectrum antibiotic and antifungal treatment, fever persisted together with diarrhea. The appearance of meningeal signs prompted a lumbar puncture that revealed a slight inflammation with six white blood cells/mm 3, but normal protein and glucose levels. Microbiological investigations revealed a positive enterovirus RT-PCR signal in the lower respiratory specimens (BAL), plasma, cerebrospinal fluid (CSF), and stools. Viral culture was positive for enterovirus in the respiratory tract and stools. Additional extensive microbiological investigations were all negative for any other bacterial, fungal, or viral infections. Disseminated enteroviral disease was diagnosed and the clinical condition improved rapidly after immunoglobulins were infused. This infusion was followed by a clearance of the infection in blood as shown by a negative RT-PCR assay at day 7 after infusion without relapse or evidence of persisting enteroviral infection. Genomic Investigations The full-length enterovirus genomes were sequenced directly from BAL, stool, plasma (at days 0 and 4) and CSF specimens (Genbank accession numbers: EU to EU414335). A whole genome BLAST search and a phylogenetic tree with available EV full-length genomes ( enterovirus/hev-a/hev-a_seqs.htm) revealed that this strain clusters with other EV71 serotypes within the human EV-A species. This serotyping was confirmed by immunofluorescence with an anti-ev71 monoclononal antibody applied on the BAL and the stool isolates grown in Vero cells. This clinical enterovirus strain is related to the genogroup C1. Its full-length polyprotein sequence was then compared to publicly available full-length EV71 sequences and linked with the identified associated clinical conditions. This large-scale inter-host analysis did not identify any genomic features that could be related to specific clinical features or to disease severity. This finding indirectly supported the completion of an intra-host full-length genome analysis to find critical residues that could promote virus dissemination and invasion of the CNS. Site-Specific Genome Analysis Genomic DNA sequences and polyprotein comparisons of the five different specimens revealed two non-synonymous substitutions at positions 662 and 1050, and one synonymous substitution at position 1906 of the EV71 polyprotein (GenBank accession number: AAB ). These positions correspond to amino acid 97 of VP1, 38 of 2B, and 175 of 3D (Table 1). None of these three mutations had any effect on the RNA secondary structure in the specific regions (data not shown). No other mutations were observed. VP1 97 leucine to arginine substitution. Compared to the initial sampling (day 0) from the lower respiratory tract, which contained a leucine at position 97 of the VP1 capsid protein (herein referred to as VP1 97L ), an arginine was present at this position (VP1 97R ) in the day 1 plasma and in the CSF sampled at day 5. In both specimens (day 1 plasma and day 5 CSF), a mixed population was not observed and only VP1 97R was present. The stool specimen sampled at day 1 contained both residues at this position, suggesting that the stool harbored a mixture of these two different species. In stool, viral culture isolated the VP1 97R subspecies as the unique and dominant strain, whereas in the respiratory specimen only the VP1 97L subspecies was isolated. This leucine to arginine substitution (L97R), located within the VP1 capsid protein, is a non-conservative change that replaces a hydrophobic non-polar residue with one that is positively charged. Based on sequence alignments to other picornavirus VP1 proteins and the comparative models of EV71 VP1 that we generated, residue 97 is located in the BC loop, a known dominant immunogenic site [24 30] situated near the putative cellular receptor binding site. An extensive alignment of 952 full-length VP1 sequences of EV71 isolates from GenBank, including our clinical specimens, revealed that this L97R substitution has previously been identified in only two other isolates, one from a meningitis case (GenBank accession number: AAB63227) and PLoS Pathogens 2 July 2012 Volume 8 Issue 7 e

113 Disseminated Human Enterovirus 71 Infection Table 1. Genome evolution at the nucleotide and amino-acid level according to the time and site of sampling. Time Site VP1 nt* VP1 aa* 2B nt 2B aa 3D nt 3D aa 1985 (290) 662 (97) 3149 (113) 1050 (38) 5718 (525) 1906 (175) Day 0 BAL T Leu T Val C Leu Day 1 Plasma 1 G Arg C Ala C Leu Day 1 Stool T + G Leu + Arg C Ala C + T Leu Day 4 Plasma 2 G Arg C Ala C Leu Day 5 CSF G Arg C Ala C Leu Positions of nucleotide and amino acid substitutions are listed in reference to both the full-length EV71 polyprotein (number on left of column) and within the affected protein (in parenthesis). BAL: bronchoalveolar lavage sample; CSF: cerebrospinal fluid sample; nt: nucleotide position; aa: amino acid position. *Positions are indicated relative to EV71 BrCr strain (GenBank accession number: U22521). Of note, real-time RT-PCR performed on plasma samples collected at days 1, 5 and 7 presented 27, 32 and 45 ct values, respectively, and immunoglobulins were injected at day 5. doi: /journal.ppat t001 another from a case with an unspecified condition (GenBank accession number: AAF13503) [31]. Of note, it has been established that the amino acid sequence of the VP1 BC loop (residues ) is an important determinant of poliovirus host adaptation [32] and that residue changes within this antigenic site show an association with mouse neurovirulence [33]. 2B 38 valine to alanine substitution. The second amino acid substitution occurred in protein 2B, known to enhance cell membrane permeability during viral infection. The neutral nonpolar valine residue at position 38 (2B 38V ) was replaced by another neutral non-polar alanine residue (2B 38A ), resulting in a conservative substitution (V38A). This 2B 38A substitution was present as the dominant species in the consensus sequence from the stool, blood, and CSF samples. Furthermore, multiple sequence alignments of 291 EV71 polyprotein sequences in Genbank revealed that most (285) circulating strains contain an alanine at this position, while the remaining sequences either contain a valine (Genbank accession numbers: ABC69251; ABW98513; ABW98514; ACB56581; ACM47545), or threonine (ABW98520). In the case of our clinical isolates, it seems that EV71 reverted to the common 2B 38A sequence. Immune Response To investigate whether these two changes could play a role in immune escape, we established quantified viral stocks in Vero cells with the BAL (VP1 97L -2B 38V ) and the stool (VP1 97R -2B 38A ) isolates, respectively. Conservation of these two substitutions after cell passage was confirmed by re-sequencing. The two isolates were tested for seroneutralization in the presence of the patient s serum (sampled at day 4) at a time when the VP1 97R substitution was already present in plasma. Neither the BAL isolate nor the stool isolate were neutralized by the patient s serum (Table 2). A negative complement fixation assay confirmed a poor antibody response against enterovirus (data not shown). Of note, in the presence of the anti-ev71 monoclonal antibody, the growth of the VP1 97R -2B 38A stool isolate was inhibited at a dilution,1:30, whereas the BAL isolate remained insensitive, thus arguing against an immune escape advantage resulting from the VP1 97R substitution. Cell Tropism of the BAL and Stool EV71 Isolates To investigate the potential implication of the mutations on tissue tropism, we then inoculated the VP1 97L -2B 38V (BAL) and VP1 97R -2B 38A (stool) isolates on three cell lines (astrocytoma [U-87 MG], neuroepithelioma [SK-N-MC], and neuroblastoma cell lines [SH-SY5Y]) previously used as references to confirm the ability of poliovirus [34,35] or EV71 [23,36] to infect cells of neural origin. Figure 1 shows that the stool isolate presents a strong replication advantage over the respiratory tract specimen in cells of neuroblastoma origin, whereas the two isolates replicate in similar fashion in the astrocytoma cell line (data not shown). Of note, neither of the two isolates was able to grow in neuroepithelioma cell lines, although a wild type poliovirus used as control grew easily under the same conditions (data not shown). Role of the VP1 97R and 2B 38A Mutations Investigated by Reverse Genetics in the Clinical Isolate Backgrounds To assess the implication of each of the two non-synonymous substitutions governing the replicative advantage of the stool isolate in cells of neuroblastoma origin, we designed four infectious clones strictly similar to the full-length sequences of the stool or the Table 2. Seroneutralization assay with the patient s serum. Clinical isolate Antibody source EV71mAB inhibitory dilution Patient serum (d4) inhibitory dilution Stool,1:30,1:5** BAL,1:10*,1:5** *smaller dilution not tested; **Patient serum was toxic for cells at a dilution,1:5. d4: patient serum sampled at day 4. BAL: bronchoalveolar lavage. doi: /journal.ppat t002 PLoS Pathogens 3 July 2012 Volume 8 Issue 7 e

114 Disseminated Human Enterovirus 71 Infection Figure 1. Replication efficiency of the lower respiratory tract (BAL) and stool isolates in Vero versus SH-SY5Y cells. Vero cells (A B) and SH-SY5Y cells (C D) were infected with equivalent concentrations of BAL and stool isolates and replication was assessed by immunofluorescence with anti-ev71 monoclonal antibody. doi: /journal.ppat g001 BAL specimens, as well as two that harbor either the VP1 97R or the 2B 38A substitutions alone (Figure 2A). We analyzed the replication efficiency of these four constructs in neuroblastoma cells (Figure 2B 2C) and other cell types (Table 3). As expected, and in contrast to the VP1 97L 2B 38V and VP1 97L 2B 38A clones that presented a delayed growth phenotype in SH-SY5Y, the virus with the stool isolate sequence and the clone with the VP1 97R substitution only grew very efficiently in Vero and SH-SY5Y cells, suggesting that the VP1 L97R substitution confers a replicative advantage in neuroblastoma cell line. The immunofluorescence results were corroborated by single-step replication analysis of the four derivatives in Vero and SH-SY5Y cell lines. Although the four derivatives demonstrated similar growth kinetics in Vero cells, the derivatives with the VP1L97R substitution (pclvp1 97R 2B 38A and pclvp1 97R 2B 38V ) presented a strong replicative advantage in SH-SY5Y cells (Figure 2C). A significant difference linked to the VP1 sequence was also observed in cells of colorectal adenocarcinoma origin (Caco-2) with the opposite phenotype in this cell line since the growth advantage was conferred by the VP1 97L sequence (Table 3). To further confirm the contribution of the VP1 L97R substitution towards the replicative advantage in cells of neuroblastoma origin, these changes were introduced individually or together in the EV71BrCr (GenBank accession number: AB ) infectious clone background (kindly provided by Prof. M Arita, National Institute of Infectious Diseases, Tokyo, Japan) that presents 81% nt and 96% aa identity with the sequence of the clinical isolates. Of note, a Lys at position 98 that reduced the replication of pbrcr in Vero cells due to the introduction of a positively charged aa was substituted with a Glu to restore normal replication. Although the replication of BrCr derivatives is very low in neuroblastoma cells, the trend was similar to that observed with the four infectious clones (data not shown). Competition Experiments Competition experiments were performed in Vero, SH-SY5Y, and Caco-2 cells to confirm these observations. For this purpose, equimolar amounts of RNA from the infectious clones derived from the stool (pclvp1 97R 2B 38A ) and BAL (pclvp1 97L 2B 38V ) isolates were co-transfected in these three cell lines. The supernatant was collected and viral sequences analysed at different time points post-transfection. As early as 24 h post-transfection, the observed dominant species in Vero and Caco-2 cells was that with VP1 97L. Regarding SH-SY5Y, at the beginning of the competition (24 h post-transfection) the population with VP1 97L appeared to slightly dominate over the VP1 97R population. However, the situation reversed after 4 days and the VP1 97R population became the dominant species (between 24 and 48 h after the first passage) (Figure 3B, left panel). The fact that the VP1 97L sequence dominates shortly after transfection suggests that once inside the cell, this position confers a replicative advantage over the VP1 97R sequence. Therefore, the VP1 97R sequence likely presents an advantage at the cell entry stage of the viral growth cycle. Interestingly, in two of four competition experiments, the VP1 97R substitution was rapidly associated with a second substitution (glutamate to glycine) present at position 167 (E167G) of VP1 (VP1 167G ) (Figure 3B, right panel). By retrospective sequence analysis of SH-SY5Y cells infected with the stool isolate or transfected with the stool or pclvp1 97R 2B 38A derivatives, position 167 was almost always mutated into a glycine or an alanine. Alignment of the 952 VP1 sequences currently available in Genbank shows that only one other EV71 sequence (GenBank accession number: AAF ) contains an alanine at position 167. Interestingly this strain also has an arginine at position 97 of VP1 (VP1 97R ). Finally, to investigate any potential implication of the L97R substitution regarding sensitivity to interferon beta, we cotransfected pclvp1 97R 2B 38A and pclvp1 97L 2B 38V in Vero cells (that do not produce, but are sensitive to interferon [37,38]) pretreated with interferon beta. The viral replication was strongly reduced by the presence of interferon beta and the VP1 97R substitution did not provide any advantage to the virus since the pcl VP1 97L 2B 38V construct was dominant after 24 h in Vero cells in the presence or absence of interferon (data not shown). VP1 Structural Modelling and Virus Binding Assay To determine if the substitution at VP1 residue 97 could have a structural impact and/or influence how the viral capsid interacts with cellular receptors or co-receptors, we generated and validated comparative models of EV71 VP1 97L and VP1 97R based on the known VP1 structures of 10 other closely related picornaviruses. Comparison of the energy signatures and structures of the models revealed that VP1 97R has no significant energetic or backbone conformational differences relative to VP1 97L (data not shown), suggesting that this substitution functions by influencing interactions at the capsid-host cell interface. To further assess this possibility, we aligned all known picornavirus VP1 structures that are in complex with their corresponding cellular receptors. In seven of the eight VP1-receptor structures (PDB accession codes listed in Figure 4A), the receptors bind in a canyon that contains the base of the BC loop, albeit in different orientations. One of the eight structures (PDB 3dpr: a human rhinovirus 2 [HRV2] bound to its receptor) revealed that the receptor interacts with VP1 not in the canyon, but directly above the BC loop at the five-fold axis of symmetry of VP1 (Figures 4A and C). We then aligned our EV71 VP1 models to the VP1 molecules in these structures and observed that EV71 VP1 97 is within Å of the receptor surfaces. Given that amino acid sequences within the BC loop contribute to receptor selectivity among picornaviruses [32,33] (Figure 4D), and that different strains interact with their receptor in different orientations and regions, we speculated that the positive charge introduced by the VP1 97R substitution could be located at the interface of human EV71 receptors and facilitate interactions with PLoS Pathogens 4 July 2012 Volume 8 Issue 7 e

115 Disseminated Human Enterovirus 71 Infection Figure 2. Schematic representation of the infectious clones derived from the stool and BAL isolates and their derivatives and assessment of their replication efficiency in Vero versus SH-SY5Y cells. (A) Representation of the infectious clones. (B C) Vero cells (I IV) and SH-SY5Y cells (V VIII) were infected with infectious clones derived from the clinical isolates. (B) Replication was assessed by immunofluorescence with an anti-ev71 monoclonal antibody. The cell type and residue at positions VP197 and 2B38 are indicated on the left and bottom of the figure, respectively. The TCID50/ml performed after 5 days on the SH-SY5Y infected cell supernatant is indicated with a bar on the left side of each panel. (C) Increase in viral RNA load as quantified by real-time RT-PCR at different time points post infection with the four derivatives. Vertical bars indicate minimum/maximum values. Key for (A) +, positive charge, 2 negative charge, o neutral; * sequence corresponding to that of the BAL samples, ** sequence corresponding to that of the stool samples. doi: /journal.ppat g002 competence is observed in favour of pcivp197r2b38a compared to pcivp197l2b38v, which supports the importance of the VP197R substitution in the receptor binding process. Of note the VP1167 position, where the compensatory E167G mutation occurred in vitro in neuroblastoma cells, lies near the interface of VP1 monomers in the capsid assembly (Figures 4B and C). Residue 167 is positioned against another negatively charged host cell receptors. Indeed, after aligning our EV71 VP1 models to the poliovirus VP1 monomers (PDB 3epf) of a complete viral capsid assembly, the arrangement of the VP1 5-mer revealed that residue 97 was close to the five-fold axis of symmetry (Figures 4B and C) in the region known to interact with host cells [39]. This model is further supported by a virus binding assay performed in Vero and SH-SY5Y cells (Figure 5). A difference in binding PLoS Pathogens 5 July 2012 Volume 8 Issue 7 e

116 Disseminated Human Enterovirus 71 Infection Table 3. Cell tropism of the different clinical isolates infectious clone derivatives. Origin of viral stock Cell lines Vero H292 Caco-2 SH-SY5Y U-87 MG Monkey kidney Lung carcinoma Colorectal adenocarcinoma Neuroblastoma Astrocytoma, glioblastoma *pclvp1 97L 2B 38V (62.17) 2.17 (60.01) (60.29) 2.6 (61.97) 7.25 (64.8) **pclvp1 97R 2B 38A (61.22) 7.7 (63.98) 4.8 (60.93) (612.41) 4.55 (61.1) pclvp1 97L 2B 38A (60.49) 5.51 (64.47) (68.98) 0.8 (60.2) 5.87 (60.18) pclvp1 97R 2B 38V (60.34) 14 (62.61) 17.6 (60.27) (64.1) 7,83 (63.43) The percent of infected cells measured by metamorph analysis is indicated with the standard deviation calculated out of two biological replicates (in parenthesis). *derives from the bronchoalveolar sequence, **derives from the stool sequence. doi: /journal.ppat t003 Figure 3. Competition between the stool and BAL infectious clone derivatives. Vero cells, Caco-2 cells (A), and SH-SY5Y cells (B) were transfected by equimolar amounts of the stool and BAL infectious clone derivatives and virus present in the cell supernatant was analysed by sequencing post-transfection and repassage at different times. Substitutions are marked by red arrows and correspond to nucleotide (nt) and amino acid (aa) positions of the EV71 VP1 coding sequence (GenBank accession number: AAB ). doi: /journal.ppat g003 PLoS Pathogens 6 July 2012 Volume 8 Issue 7 e

117 Disseminated Human Enterovirus 71 Infection Figure 4. EV71 VP1 substitution locations relative to known receptors and capsid symmetry. (A) EV71 VP1 model highlighting the BC loop (green) and positions of VP197R (red circle) and VP1167E (orange circle) relative to known receptors (gray). Eight known picornavirus VP1-receptor complexes (PDB codes along their sides) were structurally aligned to our model using the VP1 coordinates in each structure file. The distance (,12 A ) between EV71 VP1 residue 97 and receptor surfaces is marked by vertical black dotted lines (distance between VP1 residue 97 and the 3dpr receptor, also,12 A, is not marked). (B) Five EV71 VP197R model monomers arranged in capsid symmetry based on poliovirus capsid VP1 orientations (PDB 3epf). BC loops (green) and positions of residue 97 (red circles) and residue 167 (orange circles) are highlighted. (C) Side view of VP197R capsid assembly in B, rotated 80u on the plane of this page. The curvature and thickness of the capsid surface (based on PDB 3epf capsid assembly, VIPERdb) is represented as a light gray arc. (D) Sequence alignment of VP1 clinical isolates, EV71 substrain BrCr (Genbank U22521), and polivirus (PV1 (Genbank V01149), PV2 (M12197), PV3 (K01392)) surrounding EV71 VP197 and VP1167 substitutions. Index numbers refer to EV71 VP1 residue positions. doi: /journal.ppat g004 PLoS Pathogens 7 July 2012 Volume 8 Issue 7 e