This is the submitted version of the journal article: Miró, Elisenda; del Cuerpo Casas, Margarita; Rubio, Marc; [et al.]. «Whole- genome analysis to describe a human adenovirus D8 conjunctivitis outbreak in a tertiary hospital». Journal of Medical Virology, Vol. 93, Issue 8 (August 2021), p. 4840-4845. 6 pàg. DOI 10.1002/jmv.26850 This version is available at https://ddd.uab.cat/record/285194 under the terms of the license 1 Whole-genome analysis to describe a Human Adenovirus D8 keratoconjunctivitis 2 outbreak in a tertiary hospital 3 Miro E,a# Del Cuerpo M,a Rubio M,a Berengua C,a Español M,a Marin P,a Vela JI,b 4 Pomar V,c Navarro F,a Rabella Na 5 a 6 Department of Microbiology, Hospital de la Santa Creu i Sant Pau and Institut 7 d’Investigació Biomèdica Sant Pau, Barcelona, Spain. b 8 Department of Ophthalmology, Hospital de la Santa Creu i Sant Pau, Barcelona, 9 Spain. c 10 Department of Internal Medicine, Hospital de la Santa Creu i Sant Pau, Barcelona, 11 Spain. 12 13 Running title: AdV-D8 outbreak by whole-genome sequencing (WGS) 14 15 #Adress correspondence to Elisenda Miro, emiro@santpau.cat 16 Servei de Microbiologia 17 Institut d’Investigació Biomèdica Sant Pau 18 C/ Sant Quintí num. 77-79, Planta 1- Despatx 108 19 08041 Barcelona 20 Telf: +34 935537577 21 LinkedIn 22 ResearchID 1 23 Abstract (250 words) 24 Conjunctivitis is a frequent ocular disorder due to the human adenoviruses (HAdVs). 25 Only a few of the 45 types of the HAdV-D species are associated with epidemic 26 keratoconjuntivitis. One of these is HAdV-D8. 27 A nosocomial outbreak due to HAdV-D8 is rarely described because 28 keratoconjunctivitis are clinically diagnosed and treated without the need to 29 characterize the causative agent. Also, it is difficult to type it at molecular level due to 30 the tediousness of classical typing techniques. 31 In this work we describe the characterization of an outbreak for HAdV-D8 using the 32 recent whole-genome sequencing method (WGS). 33 Of the 363 patients attended between July 13 and August 13, 2018, 36 may have 34 acquired intra-hospital conjunctivitis. Eleven from 22 samples from the Virology section 35 were selected for WGS analysis. 36 WGS results showed that ten out of eleven AdV-D8 strains were closely related. The 37 remaining strain, Case 28 was more related to a public sequence obtained from an 38 outbreak that took place in Germany. WGS results showed that our HAdV-D8 strains 39 had a coefficient of similitude from 89.5% to 94.3%. 40 Whole-genome sequencing is useful in a clinical setting because it avoids viral culture 41 or specific PCR-sequencing. Sequence reads are publicly available and make it easier 42 to compare which clusters are circulating. In conclusion, whole genome sequencing 43 should play an important role in standard routine to describe viral outbreaks 2 44 INTRODUCTION 45 Conjunctivitis is a frequent ocular disorder observed in clinical practice due to a variety 46 of pathogens including bacteria, viruses or parasites. In regard to viral conjunctivitis, 47 the human adenoviruses (HAdVs) are some of the most often causal agents (1). 48 HAdVs have been associated with a wide spectrum of diseases concerning respiratory, 49 ocular, gastrointestinal, genitourinary systems and obesity (2). HAdVs are divided 50 phylogenetically into seven species, A through G. HAdV classified in B, C and E 51 species are mainly associated with respiratory diseases, those in A, D, F and G with 52 gastrointestinal disease and those in D and E with ocular diseases (2). HAdV-D is the 53 largest and most rapidly growing HAdV species, and contains viruses associated with 54 epidemic keratoconjuntivitis (EKC), a severe, hyperacute ocular surface infection that 55 usually occurs as an outbreak in schools, swimming pools, hospitals, and many other 56 locations as described in the literature(2, 3). EKC is caused by a few of the 45 types of 57 the HAdV-D species: -D8, -D37, -D53, -D56, -D54, -D64, and D-85 (4, 5). 58 The report of nosocomial outbreaks due to HAdV-D8 is rarely described because 59 keratoconjunctivitis are clinically diagnosed, and treated without the need to 60 characterize the causative agent. And, in the event, that the etiological agent was 61 detected and the outbreak was defined, it is difficult to type it at molecular level due to 62 the tediousness of classical Adenovirus typing techniques, such as neutralization 63 testing, restriction enzyme analysis (REA) or epitope sequencing. 64 Currently the classical typing of HAdVs is being replaced by sequencing techniques 65 such as whole-genome sequencing (6). To date 103 unique HAdV genotypes 66 (http://hadvwg.gmu.edu) are recognized. The genome of HAdV-D has a highly 67 conserved part, with a high GC content, and a minority part (less than 10%) 68 hypervariable that allows it to evolve into new genotypes (7). 3 69 In this work we describe the characterization of an outbreak for HAdV-D8 that 70 happened in the summer of 2018 at the Ophthalmology Department of a University 71 Hospital in Barcelona, Catalonia, using whole-genome sequencing method. 72 MATERIAL AND METHODS 73 Cases. Patients attended at the Ophthalmology Department of a University Hospital 74 (Barcelona) during August, 2018 with a clinical diagnosis of keratoconjunctivitis, this is: 75 red eyes, excessive lacrimation, foreign body sensation, photophobia, redness of the 76 bulbar conjunctiva, chemosis, petechial and subconjunctival haemorrhages. 77 Samples. Exudates from the bottom of the conjunctival sac, one for patient, collected 78 with one regular FLOQSwab (COPAN Diagnostics, Inc) and introduced in a conical 79 tube filled with 3 ml UTM medium. . 80 Microbiological methods. All conjunctival exudates were processed for adenovirus 81 antigen detection with use of direct immunofluorescence and viral culture. 82 For direct immunofluorescence, we used Light Diagnostics™ Adenovirus Antibody 83 (Merck). The specimens were spotted onto glass slides and processed by use of 84 standard techniques. The presence of viral antigen was indicated by the appearance of 85 characteristic intracellular apple-green fluorescence in nuclear, cytoplasmic or both 86 locations. The nuclear fluorescence is uniformly bright whereas cytoplasmic is was very 87 often dotted. 88 For viral culture, specimens were inoculated into each of 6 cell lines: human fibroblasts 89 (MRC5), human epithelial cells (Hep-2 and A-549), human rhabdomyosarcoma (RD), 90 rhesus monkey kidney (LLCMK2) and African green monkey kidney (VERO) cells. 91 Cultures were incubated for 2 weeks on a roller drum at 35ºC. Viruses were identified 92 on the basis of cytopathic effect in cell cultures and confirmed by staining with specific 93 fluorescein conjugated monoclonal antibodies. 4 94 In parallel a real-time PCR qualitative detection of DNA from Adenovirus in clinical 95 samples, using RealCycler ADNV (Progenie Molecular) and the SmartCycler (Cepheid) 96 was done. The DNA extraction was done using the BioRobot EZ1 and EZ1 DPS Virus 97 Kit (QIAgen). 98 Adenovirus genotyping. PCR reactions were set up in a total volume of 25 µl 99 containing 0.5 μM each oligonucleotide (AdTU7 5’-GCCACCTTCTTCCCCATGGC-3’ 100 and AdTU4’ 5’-GTAGCGTTGCCGGCCGAGAA-3’ for PCR, and AdnU-S’ 101 TTCCCCATGGCNCACAACAC and AdnU-A GCCTCGATGACGCCGCGGTG for 102 NESTED-PCR)(Yamin Li et al. 2015), 2.5 µl of 10X PCR Buffer with 2 mM MgCl2, 200 103 μM deoxynucleotide triphosphates, 1.25 U de FastStart™ Taq DNA Polymerase 104 (SigmaAldrich) and 5μl DNA (or 2 μl amplicon at Nested reaction). Amplification both 105 PCR and Nested, were carried out in a Thermal Cycler T100 (BioRad) for a total of 36 106 cycles. After an initial denaturation step at 94 °C for 10 min, each cycle consisted of 107 denaturation at 94 °C for 1 min, followed by annealing at 50ºC for 1 min and an 108 extension step at 72ºC for 2 min. Finally an extended extension at 72ºC for 7 min was 109 done. The PCR product was analysed on a 2% agarose gel and visualized with 110 Greensafe Premiun (NzyTech under ultraviolet light. 111 The PCR product was purified with the EXOSAP-IT PCR System (Thermo Fisher 112 Scientific) and sequenced using the Big Dye Terminator v3.1 Cycle Sequencing kit 113 (Applied Biosystems), with the same oligonucleotides used for PCR. The sequence 114 reaction was then purified with the AutoSeq G-50 Dye Terminator Removal kit and 115 loaded in the 3500 Series Genetic Analyser (Applied Biosystems). The sequences 116 obtained were analysed by the BioNumerics v.8.0 software (Applied Maths) and 117 introduced to the https://www.rivm.nl/mpf/typingtool/enterovirus/ to obtain the genotype. 118 Whole-Genome sequencing 5 119 Viral DNA was extracted from cell cultures with the DNeasy UltraClean Microbial Kit 120 (QIAgen, Germantown, USA) after concentration with 0.22 µm filter (Ibian 121 Technologies, Zaragoza, Spain). The DNA concentration and quality of extracted DNA 122 was determined by Qubit™dsDNA HS Assay Kit (ThermoFisher Scientific, 123 Waltham, MA, USA) and NanoDrop 1000 Spectrophotometer v3.8 (ThermoFisher 124 Scientific, Waltham, MA, USA). The libraries were prepared using a Nextera XT v.01 125 kit (Illumina Inc., San Diego, CA, USA) and sequenced on an Illumina NovaSeq 6000 126 sequencer (Illumina Inc., San Diego, CA, USA) to generate 2 x 150 bp paired-end 127 reads. The sequencing reaction was done at Novogene Europe. The obtained 128 nucleotide sequences were trimmed and analysed in BioNumerics v7.6 (Created by 129 Applied-maths NV. Available from https://www.applied-maths.com). Neighbour-joining 130 method was used for the phylogenetic tree analysis, and showed in a maximum 131 parsimony tree. 132 133 RESULTS 134 On August 14, 2018, the Ophthalmology Department have notified to Infectious 135 Diseases Unit an unexpected increase in the number of patients with suspected 136 epidemic keratoconjunctivitis (EKC), who intervened asking for a list of all patients 137 diagnosed with conjunctivitis in the last month (from 13.07 to 13.8). No case is detected 138 between the clinical staff. Exudate from the bottom of the conjunctival sac from these 139 patients was submitted to the Microbiology Department to make a description of the 140 causal agent. 141 Of the 363 patients attended at the Ophthalmology Department between July 13 and 142 August 13, 36 may have acquired intra-hospital conjunctivitis (Table 1). The first case 143 diagnosed with conjunctivitis was observed on July 20, which had previously been 144 visited on May 8, for cataracts pathology. The second case took place on August 2, but 6 145 the patient had been previously visited on July 22. Two days after the first case. Both 146 patients shared the fencing lamp. But, it is between August 5 and 8, that one peak that 147 included 11 cases make the alert of a possible outbreak (Figure 1). All cases visited 148 through July, just after case 1 shared the fencing lamp. From August 14, the Infectious 149 Diseases Unit get involved in the control of the outbreak and an new 22 cases were 150 found until August 30 (Figure 1). 151 The 22 samples, arrived at Virology section, were studied by fluorescence and viral 152 culture. In one sample (case 30) a viral culture contamination was detected. Among the 153 remaining 21 samples, 13 grew up in the viral culture for adenovirus, and they were 154 genotyped. The genotyping results showed that in all cases it was HAdV-D8. 155 In order to confirm that they were the same clone we decided to establish the clonal 156 relationship between the strains by whole-genome sequencing. Nevertheless 2 out of 157 13 strains were disregarded because the DNA extraction did not meet the requirements 158 of quantity and purity (cases 35 and 36). 159 The 11 genome sequences analyzed formed a cluster supported by a bootstrap value 160 of 100% in 1000 permutations. Figure 2 shows a Maximum Parsimony Tree to 161 represent phylogenetic relationships among our 11 studied strains together with 6 162 HAdV-D8 genomes obtained from GenBank and used as non-related strains. In this 163 case the definition of an outbreak has been simple because the isolated strains in our 164 hospital showed values of genetic distance between 100 and 700, except for one case. 165 Case 28 was genetically closer related to two of the strains that we introduced as 166 unrelated to our outbreak (AB448769 and AB448767). These two strains are closely 167 related to each otherwithin an outbreak that took place in Germany (4). 168 Genomic sequences comparison of the epidemical strains showed between 94.3% and 169 89.5% of coefficient similitude. This similitude is similar to the HAdV-D8 sequences not 170 related to our outbreak. 7 171 172 DISCUSSION 173 Keratoconjunctivitis (EKC) is a severe infectious eye disease associated with the 174 HAdV-D8, a DNA virus widely spread everywhere and widely described as the 175 causative agent of epidemic outbreaks (1). 176 On August 14, 2018, the Infectious Diseases Unit of Sant Pau Hospital was notified an 177 increase of the number of patients who were attended at the Ophtalmological 178 Department, and later developed symptoms of epidemic EKC. This fact conducted an 179 investigation, which identified 36 patients with EKC. Observations in the clinical staffs 180 found that all patients had been sharing the fencing lamp instrument, possible cause of 181 transmission.Previous studies have demonstrated that adenoviruses can persist on 182 environmental surfaces for several weeks (8). 183 Twenty-two of the 36 cases were suitable to a virological study (which included viral 184 culture with subsequent PCR and sequencing) and 13 of these were positive for HAdV- 185 D8. In order to describe and characterize the outbreak, we decided to use the whole- 186 genome sequencing method.Whole genome sequencing is technically less tedious 187 than REA (restriction enzyme analysis) (11) and has demonstrated sufficient 188 discriminatory power (2, 4, 9, 10). 189 The HAdV-D8 genome is highly conserved (3) and WGS analysis is more exhaustive 190 than the amplification and sequencing of different regions of the major capsid genes 191 (penton base, hexon, and fiber genes) used also by different authors (3). 192 In our outbreak, whole-genome sequencing with phylogenetic analysis describe a 193 monophyletic cluster of patients infected with HAdV-D8. This is potentially explained for 194 sharing a fencing lamp. The WGS results showed that the HAdV-D8 strains isolated 195 during the epidemic period had a high coefficient of similitude, but lesser than those 8 196 obtained by Hage et al. (4). In that study (4), authors found 23 samples with 99.85% of 197 identity which, were divided into two nodes. Nevertheless, our strains closer related 198 between them than with strains isolated from Germany (KP016741, KP016743), US 199 (KT340056, KT340070) or Japan (AB448767, AB448769). We selected these strains 200 from other studies as strains unrelated to the outbreak. The case 28 strain has a 201 greater distance to the HAdV outbreak strains. This greater genetic distance confirms 202 the idea that a single outlier may also circulate in parallel to the outbreak (4). 203 The outbreak was determined as closed at the end of August. After that, only three 204 cases of conjunctivitis were arrived at Ophthalmology Emergency Department, one in 205 September (HAdV-D8) , one in December (HAdVB3), as well as another one in 206 November (HAdV-D37) at Emergency Paediatrics. 207 In conclusion, whole-genome approach has shown us the utility of adenovirus 208 sequencing in a clinical setting. BOur whole-genome approach does not need viral 209 culture (although we did not do it directly from clinical samples because we had 210 previously isolated the virus strain), or specific PCR-sequencing. Also, reads are 211 publicly available making it easier to compare circulating clusters for each site, country, 212 citiey and even hospitals within the same city. In our experience, whole genome 213 sequencing should play a mandatory role in standard routine to describe viral 214 outbreaks 9 215 REFERENCES 216 (1) Balasopoulou A, Kokkinos P, Pagoulatos D, Plotas P, Makri OE, Georgakopoulos CD, 217 Vantarakis A. 2017. A molecular epidemiological analysis of adenoviruses from 218 excess conjunctivitis cases. BMC Ophthalmol 17:51. 219 (2) Ismail AM, Cui T, Dommaraju K, Singh G, Dehghan S, Seto J, Shrivastava S, 220 Fedorova NB, Gupta N, Stockwell TB, Halpin R, Madupu R, Heim A, Kajon AE, 221 Romanowski EG, Kowalski RP, Malathi J, Therese KL, Madhavan HN, Zhang Q, 222 Ferreyra LJ, Jones MS, Rajaiya J, Dyer DW, Chodosh J, Seto D. 2018. Genomic 223 analysis of a large set of currently –and historically- important human adenovirus 224 pathogen. Emerg Microb Infect. 7:10. 225 (3) Lei Z, Zhu Z, ma ci wang B, Mei H, Li H, zeng gong ga D, Jie G, ma bu chi M, Zhang 226 S, Ma C, Xu W. 2017. Outbreaks of epidemic keratoconjunctivitis caused by human 227 adenovirus type 8 in the Tibet Autonomous Region of China in 2016. PLoS ONE 228 12:e0185048. 229 (4) Hage E, Espelage W, Eckmanns T, Lamson DM, Pantó L, Ganzenmueller T, Heim A. 230 2017. Molecular phylogeny of a novel human adenovirus type 8 strain causing a 231 prolonged, multi-state keratoconjunctivitis epidemic in Germany. 232 www.nature.com/scientificreports/ 7:40680. 233 (5) Matsuura K, Terasaka Y, Uchio E, Saeki Y, Fujimoto T, Hanaoka N, Miyazaki D, Inoue 234 Y. 2019. Human adenoviral type 54 keratoconjunctivitis accompanied by stellate 235 keratitis and keratic precipitates: two cases. BMC Ophthalmol 19:7. 236 (6) Seto D, Chodosh J, Brister JR, Jones MS. 2011. Using the whole-genome sequence 237 to characterize and name human adenoviruses. J Virol 85:5701-5702. 238 (7) Ismail AM, Lee JS, Lee JY, Singh G, Dyer DW, Seto D, Chodosh J, Rajaiya J. 2018. 239 Adenoviromics: Mining the Human Adenovirus Species D Genome. Front Microbiol, 240 9:2178. 10 241 (8) Killerby ME, Stuckey MJ, Guendel I, Sakthivel S, Lu X, Erdman DD, Schneider E, 242 Fagan R, Davis MS, Watson JT, Gerber SI, Biggs HM Ellis EM. 2017. Epidemic 243 Keratoconjunctivitis outbreak associated with Human Adenovirus Type 8 – U.S. 244 Virgin Islands, June-November 2016. MMWR Morb Mortal Wkly Rep, 66:811- 245 812. 246 (9) Houldcroft CJ, Roy S, Morfopoulou S, Margetts BK, Depledge DP, Cudini J, Shah D, 247 Brown JR, Romero EY, Williams R, Cloutman-Green E, Standing JF, Hartley JC, 248 Breuer J. 2018. Use of whole-genome sequencing of Adenovirus in immunocomprised 249 pediatric patients to identify nosocomial transmission and mixed-genotype infection. J 250 Infect Dis, 218:1261-1271. 251 (10) Lamson DM, Kajon AE, Shudt M, Quinn M, Newman A, Whitehouse J, Greenko J, 252 Adams E, St. George K. 2018. Molecular typing and whole genome next generation 253 sequencing of human adenovirus 8 strains recovered from four 2012 outbreaks of 254 keratoconjunctivitis in New York State. J Med Virol. 90:1471-1477. 255 (11) Adhikary AK, Ushijima H, Fujimoto T. 2012. Human adenovirus type 8 genome typing. 256 J Med Microbiol. 61:1491-1503. 257 ADDITIONAL INFORMATION 258 Competing financial interest: The authors declare no competing financial interest. 259 Accession Codes: The HAdV-D8 genomic sequences produced in this study have 260 been deposited under the following Bio-Project number PRJNA669467. 11 261 Table 1. Epidemiological data from the 36 cases possible implicated in the epidemical adenoviral conjunctivitis. Case Age Gende Previously Visits Diagnostic at previous visits Data diagnosis Adenovirus Treatment r conjuctivitis detection 1 63 Male 08.05 20.07 Not performed Tobradex, Aquoral Forte, SF 2 54 Male 22.07 Irritative conjunctivitis 02.08 Not performed Hyabak, SF 3 65 Female 05.07/09.07/31.07 05.08 Not performed Aquoral Forte, Isoptoflucon 4 37 Male 01.06/28.07 Corneal erosion by contact lenses 06.08 Not performed Tobradex, Thealoz, 5 52 Female 11.06/28.07 Retinal detachment / corneal erosion 06.08 Not performed Thealoz, SF 6 61 Male 18.07 Retinal detachment 06.08 Not performed Aquoral Forte, SF 7 76 Female 27.07/01.08 06.08 Not performed DXM, SF 8 33 Male 28.07 Chemical conjunctivitis 06.08 Not performed Tobradex 9 33 Female 29-31.07/01.08 Corneal erosion 06.08 Not performed Thealoz, SF 10 73 Female 03.08 Bacterial conjunctivitis 06.08 Not performed Hyabak, Icol pomada 11 55 Male 16.07/30.07 Retinal detachment 08.08 Negative De Icol, Neovis 12 82 Male 30.07 08.08 Not performed Thealoz, SF 13 72 Female 02.08 Triassic tab 06.08 Not performed Icol pomada, SF 12 14 53 Female 04/06.08 Conjunctivitis CE 14.08 Not performed Thealoz, SF 15 35 Male 06.08 Retinal control 14.08 Not performed De Icol, SF 16 67 Female 07.08 Control of cryocystectomy 17.08 Negative Xilin, SF 17 73 Male 31.07 Corneal lesions 17.08 Negative Thealoz, De Icol 18 55 Female 06.08 Granulomatous anterior uveitis 17.08 Positive Belcils, SF 19 81 Female 06.08 Glaucoma control 17.08 Positive De Icol, Hyabak, SF, 20 79 Female 16.07 Retinal control 20.08 Negative Xilin, SF 21 59 Male 01.08 Vitrectomy control 20.08 Negative Aquoral Forte, SF 22 87 Female 6.08/14.08 Ophthalmia paresia IIIpc 20.08 Positive 23 70 Male 08,08/16.08 Retinal control 20.08 Negative Aquoral Forte, Tobradex, SF 24 24 Female 13.08 Irritative conjunctivitis 22.08 Positive Aquoral Forte, SF 25 60 Female 14.08 Myopia magna, control 22.08 Positive Hyabak, SF 26 65 Female 19.08/21.08 QPS 22.08 Negative Aquoral, SF 27 58 Male 13.08 Retinal detachment, control 23.08 Positive Neovis, SF 28 84 Male 08.08 eyelid edema 27.08 Positive De Icol, Thealoz 29 89 Female 14.08/17.08/24.08 corneal erosion 27.08 Positive De Icol, Azarga, Thealoz 13 30 69 Male 14.08 28.08 Positive De Icol, Hyabak, SF 31 81 Female 14.08 28.08 Positive Neovis, SF 32 56 Male 14.08/21.08 desprendiment vitri/control 29.08 Negative De Icol, SF 33 74 Male 16.08 iridiotomy 30.08 Positive Hyabak, SF 34 65 Female 16.08/23.08 iridiotomy, control 30.08 Positive Thealoz, SF 35 69 Male 20.08 biocular druses 30.08 Positive Hyabak, SF 36 78 Male 20.08 Retinal detachment, control 30.08 Positive De Icol 262 263 14 264 265 266 267 268 269 270 271 272 273 Figure 1. Chronological distribution of the 36 cases with keratoconjunctivitis for AdV-D8 since their first visit to Ophtalmology department and 274 the consecutive ones. 15 275 276 Figure 2. Molecular Phylogenetic analysis by Maximum Parsimony method. The tree 277 was obtained using BioNumerics v.7.6. Bootstrap values (with 1000 permutations) in all 278 cases were 100%. 16