Human adenovirus type 8 genome typing

Correspondence A. K. Adhikary arun_adhikary@msn.com Department of Microbiology, BGC Trust Medical College, Chandanaish, Chittagong, Bangladesh Division of Microbiology, Nihon University School of Medicine, 30-1 Oyaguchi-Kamicho, Itabashi-ku, Tokyo 173-8610, Japan Department of Developmental Medical Sciences, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan Infectious Diseases Surveillance Center, National Institute of Infectious Diseases, Tokyo 162-8640, Japan


Introduction
Human adenovirus type 8 (HAdV-8) is frequently associated with epidemic keratoconjunctivitis (EKC), a severe and highly contagious eye disease (Shenk, 1996;Paval-Langston, 1994).Being a non-enveloped virus, HAdV-8 is remarkably resistant to physical and chemical agents and remains infectious at room temperature for prolonged periods, thereby facilitating transmission through fomites, ophthalmic instruments and even ocular drops (Hamada et al., 2008;Saitoh-Inagawa et al., 1999).This endows the virus with the potential to spread in the community as well as in medical facilities.
Conventional typing with a neutralization test (NT) cannot differentiate genetic diversity among the field strains of HAdV-8 (Hierholzer et al., 1991;Wadell et al., 1980).Genome typing by restriction endonuclease (RE) cleavage pattern analysis of isolates is the method of choice for elucidating this diversity.This method enables prediction of a possible correlation between genomic variation and virulence of the strain.So far, HAdV-8A-8K and HAdV-8/ D1-D12 have been described using the Asian and European classification systems of genome typing, respectively (Fujii et al., 1983(Fujii et al., , 1984;;Fujii & Yamashita, 1984;Ishii et al., 1987;Sawada et al., 1987;Sheu et al., 1987;Guo et al., 1988;Chang et al., 2001;Adhikary et al., 2003Adhikary et al., , 2011;;Ohguchi et al., 2003;Jin et al., 2011;Taka ´cs et al., 1983;Adrian et al., 1990;de Jong et al., 1992;Tanaka et al., 2000).Nevertheless, there have also been several problems with genome typing, such as the application of variable numbers and types of REs as well as difficulties in the interpretation of results due to the lack of a published catalogue of HAdV-8 genome types.The aim of this review is to develop an updated organized approach to HAdV-8 genome typing by compiling all the published information.The resulting data are expected to provide support for future epidemiological studies.

Historical overview
During the summer of 1941, a very large outbreak (more than 10 000 cases) of keratoconjunctivitis occurred in Hawaii, at the naval shipyards of Pearl Harbor.Soon thereafter, it spread to the West Coast of the United States and seriously hampered wartime industrial production.1975-1978, Sapporo HAdV-8A 1976-1981, Sapporo HAdV-8B Fujii et al. (1984Fujii et al. ( ) 1980Fujii et al. ( -1981 Several thousand cases were diagnosed and most of them were in shipyard workers.The disease was thus named 'shipyard eye' (Jawetz, 1959).Subsequently, due to its epidemic nature, the illness was called 'epidemic keratoconjunctivitis' (EKC) (Hogan & Crawford, 1942).The disease was characterized by severe bilateral conjunctivitis with a painful and distressing keratitis, severe photophobia and corneal opacities with variable degree of visual impairment for months to years (Jawetz et al., 1955a(Jawetz et al., , 1955b(Jawetz et al., , 1957)).EKC was also described among German workers in the late 19th century.However, the aetiological agent remained unidentified until December 1954, when a seaman travelling from an East Asian port came to an eye clinic of the University of California for treatment of severe bilateral conjunctivitis (red eye).A conjunctival scraping was collected from the seaman (whose surname was Trim) and the causative virus of 'shipyard eye' was isolated for the first time in HeLa cell culture.The virus was typed as HAdV-8 by homologous antiserum and designated the prototype strain Trim (Jawetz et al., 1955a(Jawetz et al., , b, 1957)).Since its original isolation, HAdV-8 remains the major agent of EKC around the world and has been isolated from many outbreaks in community, industrial and medical settings, including the offices of ophthalmologists (Guyer et al., 1975;Laibson et al., 1968;Sprague et al., 1973;Dawson & Darrell, 1963;Patrick & Matthews, 1981;Buehler et al., 1984;D'Angelo et al., 1981).However, the intraserotypic genetic variability of HAdV-8 field isolates was not examined until 1983, when the viral DNA was digested by multiple REs and genetically different strains were described as genome types.Two genome type classification systems were developed separately in Asia and Europe.

Genome type designation based on numerical order: European classification
A numerical code to denominate HAdV-8 genome types was proposed by Adrian et al. (1990).The formula for designating a genome type included five pieces of information as follows: number of serotype/number of variant/ designation of first identified strain of variant/geographical origin of this strain/year of isolation of this strain.Adrian et al. (1990) used six RE patterns in alphabetical order (BamHI, BglII, HindIII, KpnI, SalI and SmaI).Later, BglI, BstEII and SacI were included in successive studies (de Jong et al., 1992;Tanaka et al., 2000).The cleavage pattern of the prototypes for all REs was labelled as enzyme code 1.
Early HAdV-8 genome type classification was begun in Europe using the numerical codes of restriction enzymes to denominate the adenovirus genome type in chronological order of the isolates (Adrian et al., 1990).Six REs (BamHI, BglII, HindIII, KpnI, SalI and SmaI) were applied to 60 strains of HAdV-8 that were isolated from Germany, Japan, Austria and Hungary over a 20-year period from 1961 to 1982.Fifty-four strains were isolated in Germany and 6 from other countries.Forty-five of these isolates were from the same source and classified as genome type HAdV-8/D1, which was similar to the Koja strain originally isolated in Japan in 1961.The prototype strain Trim had been included under HAdV-8/D1.Five genome types (HAdV-8/D2-D6) were found among 15 strains isolated over a period of 17 years from Japan, Hungary, Austria and Germany.In this study, Adrian et al. (1990) created a restriction site map of HAdV-8/D1 with BamHI, BglII, HindIII and SalI and determined the molecular mass of the fragment.They also calculated the percentage of pairwise co-migrating restriction fragments (PCRFs) in order to measure relatedness among the genome types.

Identification of HAdV-8
The conventional method of HAdV identification consists of culturing the virus and typing by NT.In the last few years, many new recombinant types of HAdVs have been identified.Some of them share the major capsid protein of HAdV-8.Therefore, instead of NT, a combination of PCR and sequencing of the hexon and fibre genes of HAdV-8 directly from a clinical sample or from isolates followed by phylogenetic analysis is recommended for molecular typing of HAdVs (Ishiko et al., 2008;Ishiko & Aoki, 2009;Kaneko et al., 2011).

Propagation of virus and DNA extraction
Most of the strains of HAdV-8 grow well in cells of human origin.rounding and swelling of cells with nuclear enlargement followed by cellular detachment from the culture surface into grape-like clusters.However, cytological changes are not always clear by light microscopy in the case of prototype strain Trim and other fastidious strains of HAdV-8 (Guo et al., 1988;Hanna & Jazetz, 1962;Wigand et al., 1983;Wigand, 1987;Golden & McKee, 1970).

DNA RE analysis
This technique requires fairly large amounts of purified or partially purified DNA, and owing to small differences in the DNA of various isolates, multiple REs are required for genome typing of HAdV-8.Usually, 0.5-1.0mg DNA is incubated with 5-10 units of RE in a 10-20 ml reaction mixture at an appropriate temperature (recommended for each RE) for 2-3 h.In addition to the clinical isolate, a reference strain is included.Reactions are stopped by addition of EDTA to a 5 mM final concentration.
The fragmented DNAs and molecular mass marker are loaded in separate wells on a horizontal submerged agarose gel and electrophoresis is done in TAE or TBE buffer.High grade and higher concentrations of gel (up to 1.2 %) yield better resolution of low molecular mass fragments (Tanaka et al., 2000).DNA in the gel can be stained with ethidium bromide (1 mg ml 21 ) after electrophoresis or the dye can be incorporated into the agarose gel during preparation.A photograph is taken under UV light (Fig. 2).
Restriction pattern analysis with schematic DNA restriction profiles and enzyme codes Fig. 3 and Fig. 4 show a comprehensive schematic restriction profile of all genome types included in the Asian and European classification systems.The profile was created based on the published restriction patterns of HAdV-8.Enzyme codes were included in the Asian classification system in accordance with the alphabetical order of the genome types as they were described over time.The cleavage pattern of the prototype for each enzyme (BamHI, HindIII, PstI, SacI, SalI and SmaI) is labelled enzyme code 1. Patterns deviating from those of the prototype are labelled 2, 3, etc.Comparison of the restriction patterns of the isolates with the schematic restriction pattern could be helpful to determine the code for the respective enzyme.

Assignment of genome type
Visual comparison of the resulting restriction patterns with the schematic restriction patterns and determination of a code for each enzyme are essential for genome typing.The distances between different digested fragments are dependent on the concentration of gel.Use of a known genome type as a reference could be helpful for comparing the results of RE analyses.If the restriction pattern of the isolate matches the schematic restriction pattern of HAdV-8, then the isolate The code is to be arranged in alphabetical order of the REs as shown in the enzyme code table (Tables 4 and 5).The genome type can then be identified by comparison of the resulting enzyme codes with the enzyme code table (e.g.HAdV-8B is determined by the enzyme code arrangement 2,2,2,2,3,2 for BamHI, HindIII, PstI, SacI, SalI, SmaI, respectively).If the codes for all enzymes are already present but the arrangement (order) is different from that for previously known genome types (e.g.HAdV-8D, HAdV-8E, HAdV-8/D12), the strain can be designated a new genome type.Again, if there is a new restriction pattern for one or more enzymes, the strain will also be a new genome type.
Fig. 5 shows the schematic pathway of the genome typing from a conjunctival swab.

Discussion
EKC is not a blinding ocular infection, yet it creates distress and ocular opacity that culminates in prolonged morbidity with possible economic impact (Barnard et al., 1973).Every year, HAdV-8 accounts for the highest numbers of EKC cases worldwide (Fox et al., 1977;Schmitz et al., 1983;Yamadera et al., 1995;Kaneko et al., 2011).NT is unable to differentiate one field strain from another because the viral epitopes taking part in this reaction are encoded by a small portion of the viral genome (Wadell, 1984).In contrast, RE analysis is a sensitive method for comparison of closely related genomes, second only to nucleotide sequence analysis (Wadell et al., 1980).Recently, whole genome sequences became available for HAdVs, but the reported numbers of strains are still limited.
In addition, whole genome sequencing is expensive in comparison to RE analysis.Both globally dispersed and geographically restricted genome types of HAdV-8 have been identified.For example, HAdV-8D/11 and HAdV-8D/ 12 are constrained to Brazil and HAdV-8G and HAdV-8H to Taiwan.However, HAdV-8P and HAdV-8E are typical examples of globally dispersed strains, as they have been circulating in many countries.Co-circulation and chronological changes of genome types have also been reported in many studies.In Hiroshima, Japan, HAdV-8A and HAdV-8B co-circulated in 1983, and from 1984 to 1988, HAdV-8E co-circulated with HAdV-8A and HAdV-8B.Afterwards, HAdV-8E circulated as a single genome type until 1995 (Adhikary et al., 2003).Some of these genome types seem to be associated with more severe clinical manifestations than others (Ishii et al., 1987;Sheu et al., 1987;Chang et al., 2001).Therefore, genome typing of HAdV-8 is indispensable in epidemiological studies that offer a chance to understand the virulent nature of different strains and to follow their circulation across different geographical regions and periods of time.
RE analysis is a labour-intensive method requiring nearly 2 weeks to complete.The application of this method is challenged by many inherent problems, including (a) the fastidious nature of HAdV-8 prototype strain Trim; (b) the use of NT as a conventional method for typing of HAdV-8, which may misidentify some of the recombinant adenoviruses as HAdV-8 due to cross-reaction; (c) lack of a schematic restriction profile for all genome types collectively; with change of cell lines or with a change in temperature, serum or other growth conditions (Guo et al., 1988;Hanna & Jazetz, 1962;Wigand et al., 1983;Golden & McKee, 1970).Moreover, repeated passage of strain Trim to obtain a sufficient amount of DNA leads to mutation of the viral genome.These properties limit the use of strain Trim as a standard in genome typing studies.HAdV-8E is included under HAdV-8/D1 and shows good growth ability (Adrian et al., 1990).This genome type could be an alternative to strain Trim in RE analysis as it is widely distributed and has been circulating over a prolonged period (198322001).
Enzyme codes were not in use in Asian classification.However, the codes for enzymes in alphabetical order and tabulated form are useful for stepwise identification of genome types.The enzyme code table helps to display the genetic variation between different genome types without  The DNA variants are denoted by numbers which are given chronologically according to the description in the literature.D1 is, by definition, the DNA variant to which prototype strain Trim belongs.The DNA restriction profiles of the prototype strain Trim are designated number 1 for each enzyme.Other profiles are consecutively numbered in chronological order of appearance of the corresponding DNA variant.
viewing restriction patterns, since the code for an enzyme is dependent on the restriction pattern.Therefore, the enzyme codes have been included in the Asian classification based on the alphabetical order of the genome types.
The percentage of PCRFs can be used to measure the genetic relatedness between HAdVs.The percentage of PCRFs is more than 50 % between the members of a species and less than 20 % between the members of different species (Adrian et al., 1986).The percentage of PCRF is 90-98 % among the HAdV-8 genome types in Asian studies, which is higher than that observed in European studies (72 % or more) (Adhikary et al., 2011;Adrian et al., 1990).HAdV-54 shares a high percentage of PCRFs with other HAdV-8 genome types.Therefore, we assume that the percentage of PCRFs is useful only when the identity of HAdV-8 has already been confirmed by the molecular method.
In analysing the problems of HAdV-8 genome typing, it is evident that a simple step-by-step procedure for HAdV-8 genome typing will require (1) correct identification of HAdV-8 by molecular methods before RE analysis, (2) a schematic restriction profile for all genome types and (3) use of enzyme codes in both classification systems.In this review, we have presented a schematic restriction profile for all HAdV-8 genome types incorporated so far in both the Asian and European classification systems by compiling all the published reports.At the same time, restriction enzyme codes for each genome type are included in the Asian classification system.We have also described a simple and organized pathway of HAdV-8 genome typing that begins with type identification and ends with interpretation of the results of RE analysis.In spite of a number of limitations, RE analysis of HAdV-8 is highly reproducible and very accurate for determining the genetic relatedness among different strains.The DNA cleavage patterns can be used as strain markers in epidemiological studies such as those evaluating nosocomial outbreaks caused by HAdV-8 (Chastel et al., 1988).

Conclusions
Despite the clinical-epidemiological impact of EKC, the number of published HAdV-8 genome typing studies is lower than expected.The lack of a schematic restriction profile of all genome types as well as the need for an organized step-by-step approach have been the major obstacles to routine use of this technique.The updated system of genome typing of HAdV-8 presented here will assist public heath institutes and research laboratories to identify either existing or novel genome types so that they can be usefully compared with clinical and epidemiological data.
Enzyme used in the study; x, enzyme not used in the study.

Fig. 4 .
Fig. 4. Schematic presentation of RE BamHI, BglI, BglII, BstEII, HindIII, KpnI, SacI, SalI and SmaI cleavage patterns of HAdV-8/ D1-HAdV-8/D12.M indicates the molecular mass standard.The DNA restriction profiles of the prototype strain Trim are designated number 1 for each enzyme.Other profiles are consecutively numbered in chronological order of appearance of the corresponding DNA variant.

Table 2 .
Adrian et al. (1990)enome type circulation: classification introduced byAdrian et al. (1990) Various continuous epithelial cell lines, such as HEp-2, HeLa, KB and A549, are useful for virus isolation, but the A549 cell line is very effective for propagation of HAdVs including HAdV-8.The cytopathic effect (CPE) of HAdV-8 in monolayer cell culture is characterized by Downloaded from www.microbiologyresearch.orgby IP: 54.70.40.11On: Thu, 25 Oct 2018 20:56:51 Downloaded from www.microbiologyresearch.orgby IP: 54.70.40.11On: Thu, 25 Oct 2018 20:56:51 . The modified Hirt's method of DNA extraction requires approximately 24 h and exploits the differences between low molecular mass viral DNA and high molecular mass cellular DNA.When 75-100 % CPE is visible, infected cells are dislodged and pelleted by low-speed centrifugation and cleaned with PBS.The viruses are released from the cells by mixing with lysis solution [10 mM Tris/HCl (pH 7.4), 10 mM EDTA, 1 % SDS], then the proteins in suspension are degraded by addition of proteinase K to a final concentration of 200 mg ml 21 at 37 u C for 1 h.After the incubation, high molecular mass cellular DNA is precipitated by addition of NaCl (5 M) to a final concentration of 1 M, and further incubated at 4 u C overnight.The suspension is then centrifuged at 15 000 g for 30 min to keep low molecular mass viral DNA in the supernatant.
The supernatant is incubated with 30 mg RNase A for 1 h to remove cellular RNA and extracted twice with phenol/ chloroform to separate DNA in the supernatant (aqueous phase).The supernatant is mixed in two volumes of absolute ethanol at 220 u C. Since DNA is insoluble in alcohols, it will aggregate together, giving a pellet upon centrifugation.After drying, DNA is suspended in TE buffer [10 mM Tris/HCl (pH 7.4), 10 mM EDTA] and quantified spectrophotometrically.

Table 3 .
Pairwise comparison of co-migrating RE cleavage fragments from the DNA of HAdV-8 genome types

Table 4 .
Human adenovirus 8 genome types: Asian classification BamHI, HindIII, PstI, SacI, SalI, SmaI.The restriction patterns of the HAdV-8 prototype (HAdV-8P) for each enzyme are designated number 1.The other patterns are consecutively numbered in alphabetical order of genome types.
*Enzyme codes are displayed in alphabetical order:

Table 5 .
Human adenovirus type 8 genome types: European classification