A4-Year-OldMaleWhoHas a Persistent, Severe Headache · hormone replacement therapy. Eighty percent...
Transcript of A4-Year-OldMaleWhoHas a Persistent, Severe Headache · hormone replacement therapy. Eighty percent...
JOURNAL OF CLINICAL MICROBIOLOGY, Oct. 2008, p. 3259–3269 Vol. 46, No. 100095-1137/08/$08.00�0 doi:10.1128/JCM.02354-07Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Epidemic Keratoconjunctivitis Due to the NovelHexon-Chimeric-Intermediate 22,37/H8
Human Adenovirus�
Koki Aoki,1 Hiroaki Ishiko,2* Tsunetada Konno,2 Yasushi Shimada,2 Akio Hayashi,2Hisatoshi Kaneko,3 Takeshi Ohguchi,1 Yoshitsugu Tagawa,1
Shigeaki Ohno,1 and Shudo Yamazaki4
Department of Ophthalmology and Visual Sciences, Hokkaido University Graduate School of Medicine, Sapporo 060-8638, Japan1;Host Defense Laboratory, Mitsubishi Chemical Medience Corporation, Tokyo 174-8535, Japan2; Department of Microbiology,
Fukushima Medical University School of Medicine, Fukushima 960-1295, Japan3; andAIDS Vaccine Development Association, Tokyo 169-0075, Japan4
Received 8 December 2007/Returned for modification 5 March 2008/Accepted 22 July 2008
In a 2-month period in 2003, we encountered an outbreak of epidemic keratoconjunctivitis (EKC) in Japan.We detected 67 human adenoviruses (HAdVs) by PCR from eye swabs of patients with EKC at five eye clinicsin different parts of Japan. Forty-one of the 67 HAdV DNAs from the swabs were identified as HAdV-37 byphylogenetic analysis using a partial hexon gene sequence. When the restriction patterns of these viral genomeswere compared with that of the HAdV-37 prototype strain, one isolate showed a never-before-seen restrictionpattern. Within 1 year, we encountered three more EKC cases caused by a genetically identical virus: twonosocomial infections at two different university hospitals and a sporadic infection at an eye clinic. Wedetermined the nucleotide sequences of the full-length hexon and fiber genes of these isolates and comparedthem to those of the 51 prototype strains. Surprisingly, the sequence of the hexon (� determinant) loop-1 and-2 regions showed the highest nucleotide identity with HAdV-22, a rare EKC isolate. However, the nucleotidesequence of the fiber gene was identical to that of the HAdV-8 prototype strain. 22 We propose that this virusis a new hexon-chimeric intermediate HAdV-22,37/H8, and may be an etiological agent of EKC.
Human adenoviruses (HAdVs) belong to the genus Mast-adenovirus of the family Adenoviridae and infect billions ofpeople worldwide, causing various diseases, including conjunc-tivitis, respiratory infectious disease, diarrhea in infants andyoung children, and hemorrhagic cystitis (31). Currently, 51serotypes have been identified and reclassified into six species,HAdV-A to -F, on the basis of nucleotide (nt) and deducedamino acid sequences (8). Conjunctivitis due to AdV is causedmainly by HAdV-3 (in HAdV-B), -4 (in HAdV-E), -8, -19, and-37 (all in HAdV-D). Among these, HAdV-8, -19, and -37cause more-severe epidemic keratoconjunctivitis (EKC) thanthe others (2–4). Serotyping of HAdVs is generally performedby virus isolation followed by a neutralization test (NT) andhemagglutination inhibition test (HI) using type-specific anti-serum (29). To date, amplification of the genome by PCR anddetermination of the nucleotide sequences followed by phylo-genetic analysis have become the general procedure for boththe classification and identification of HAdVs (21, 22, 28). In2003, we encountered an EKC outbreak in Japan (InfectiousAgents Surveillance Report [Weekly reports of adenovirus isola-tion/detection from epidemic keratoconjunctivitis cases, 2001–2007; http://idsc.nih.go.jp/iasr/prompt/graph/ad3.gif]) and de-tected HAdV DNA in 67 swabs from EKC patients who visitedfive eye clinics in different parts of Japan. Among the 67
HAdVs, 41 were identified as HAdV-37 by a phylogeneticanalysis based on the partial sequence of conserved region 4(C4) in the hexon gene (22). When the restriction patterns ofviral genomes of 10 representative isolates from the swabswere compared with that of the HAdV-37 prototype strain,one strain, C075/Matsuyama/2003, showed a different restric-tion pattern. This virus was isolated in Matsuyama city onShikoku Island. During the same period, we observed a noso-comial infection due to genetically the same virus as C075/Matsuyama/2003 at a university hospital in Tokyo. In the fol-lowing year (2004), we encountered nosocomial EKC due toHAdV-37 in six hospitals and HAdV-19a in one hospital. Wealso found one nosocomial infection caused by a virus genet-ically identical to C075/Matsuyama/2003 at a university hospi-tal in Yamaguchi. A C075/Matsuyama/2003-like virus alsocaused a sporadic infection at an eye clinic in Fukui, in thewestern part of Japan in 2004.
In this study we determined the nucleotide sequences of theentire hexon and fiber genes of the C075/Matsuyama/2003-likeisolates and compared them with those of the 51 prototypestrains. The 350 bp of the partial hexon C4 sequence showed99.7 to 100% identity with that of the HAdV-37 prototypestrain, whereas hexon loops 1 (L-1) and 2 (L-2), which containthe NT epitope, showed 99.6 to 100% nucleotide identity toHAdV-22. Moreover, the fiber knob, responsible for cell tro-pism, has a nucleotide sequence identical with that of theHAdV-8 prototype strain, although the penton base has anucleotide sequence identical to that of HAdV-37. Therefore,we propose that this virus is a novel hexon-chimeric interme-diate HAdV containing hexon C4, hexon L-1, hexon L-2, fiber
* Corresponding author. Mailing address: Host Defense Laboratory,Mitsubishi Chemical Medience Corporation, Shimura 3-30-1, Itabashi-ku, Tokyo 174-8555, Japan. Phone: 81-3-5994-2340. Fax: 81-3-5994-2972. E-mail: [email protected].
� Published ahead of print on 13 August 2008.
3259
TA
BL
E1.
Typ
eid
entifi
catio
nof
HA
dVD
NA
dete
cted
inco
njun
ctiv
alsw
abs
from
patie
nts
with
EK
Cby
phyl
ogen
etic
anal
ysis
Rep
rese
ntat
ive
isol
ate/
orig
in(i
nJa
pan)
/yr
Sour
ceT
ypin
gby
phyl
ogen
etic
anal
ysis
No.
ofis
olat
esor
swab
s
No.
ofD
NA
copi
es/m
lH
ighe
st-
scor
ing
prot
otyp
e(s)
ofth
epa
rtia
lhe
xon
regi
on(3
50bp
)
Iden
tity,
%
Nex
t-hi
ghes
t-sc
orin
gpr
otot
ype(
s)of
the
part
ial
hexo
nre
gion
(350
bp)
Iden
tity,
%G
enom
ety
peG
enB
ank
acce
ssio
nno
.M
inim
umM
axim
umM
ean
C00
4/Sa
ppor
o/20
03Sw
aba
HA
dV-3
79
4.8
�10
21.
5�
107
2.0
�10
6H
AdV
-37
100.
0H
AdV
-25
98.9
HA
dV-3
7vA
B43
7260
C00
3/Sa
ppor
o/20
03Sw
aba
HA
dV-4
37.
2�
104
3.3
�10
71.
1�
107
HA
dV-4
96.7
HA
dV-2
5,-3
2,-3
7,-4
989
.3N
Tc
AB
4372
61
C04
6/T
okyo
/200
3Sw
aba
HA
dV-3
71
1.2
�10
31.
2�
103
1.2
�10
3H
AdV
-37
100.
0H
AdV
-25
98.9
HA
dV-3
7vA
B43
7262
C00
1/T
okyo
/200
3Sw
aba
HA
dV-1
9a2
3.4
�10
67.
2�
107
3.8
�10
7H
AdV
-25,
-26,
-38,
-49
97.4
HA
dV-2
4,-3
2,-4
297
.1H
AdV
-19a
AB
4372
63
C10
1/T
okyo
/200
3Sw
aba
HA
dV-B
12.
8�
106
2.8
�10
62.
8�
106
HA
dV-1
499
.7H
AdV
-11
98.7
NT
AB
4372
64C
075/
Mat
suya
ma/
2003
Swab
aH
AdV
-37
13.
3�
106
3.3
�10
63.
3�
106
HA
dV-3
799
.7H
AdV
-24,
-25
98.6
Non
-HA
dV-3
7A
B36
9371
C01
8/M
atsu
yam
a/20
03Sw
aba
HA
dV-3
77
2.5
�10
52.
0�
108
3.7
�10
7H
AdV
-37
100.
0H
AdV
-25
98.9
HA
dV-3
7vA
B43
7265
C02
9/M
atsu
yam
a/20
03Sw
aba
Nov
elA
dV12
1.3
�10
53.
4�
107
6.3
�10
6H
AdV
-8,-
29,
-38,
-43,
-46
95.4
HA
dV-2
2,-2
4,-2
5,-3
0,-3
2,-3
3,-3
7,-4
5,-4
7
95.1
Nov
elA
dVA
B43
7266
C06
4/M
atsu
yam
a/20
03Sw
aba
HA
dV-B
21.
4�
106
7.6
�10
73.
9�
107
HA
dV-3
98.3
HA
dV-7
97.3
NT
AB
4372
67C
011/
Kum
amot
o/20
03Sw
aba
HA
dV-3
720
1.3
�10
41.
3�
108
1.8
�10
7H
AdV
-37
100.
0H
AdV
-25
98.9
HA
dV-3
7vA
B43
7268
C04
9/K
umam
oto/
2003
Swab
aH
AdV
-19a
13.
6�
107
3.6
�10
73.
6�
107
HA
dV-2
5,-2
6,-3
8,-4
997
.4H
AdV
-24,
-32
,-42
97.1
HA
dV-1
9aA
B43
7269
C08
3/It
oman
/200
3Sw
aba
HA
dV-3
73
3.0
�10
41.
3�
107
4.9
�10
6H
AdV
-37
99.7
HA
dV-2
598
.3H
AdV
-37v
AB
4372
70C
060/
Itom
an/2
003
Swab
aH
AdV
-19a
16.
0�
107
6.0
�10
76.
0�
107
HA
dV-2
5,-2
6,-3
8,-4
997
.4H
AdV
-24,
-32
,-42
97.1
HA
dV-1
9aA
B43
7271
C03
9/It
oman
/200
3Sw
aba
HA
dV-4
16.
4�
106
6.4
�10
66.
4�
106
HA
dV-4
96.7
HA
dV-2
5,-3
2,-3
7,-4
989
.3N
TA
B43
7272
C08
5/It
oman
/200
3Sw
aba
Nov
elA
dV1
5.2
�10
65.
2�
106
5.2
�10
6H
AdV
-8,-
29,
-38,
-43,
-46
95.4
HA
dV-2
2,-2
4,-2
5,-3
0,-3
2,-3
3,-3
7,-4
5,-4
7
95.1
Nov
elA
dVA
B43
7273
C05
8/It
oman
/200
3Sw
aba
HA
dV-B
21.
3�
107
5.6
�10
73.
5�
107
HA
dV-3
99.0
HA
dV-7
98.0
NT
AB
4372
748/
Tok
yo/2
003
Swab
bH
AdV
-37
92.
0�
102
3.1
�10
10
4.2
�10
9H
AdV
-37
100.
0H
AdV
-25
98.9
NT
AB
4372
751/
Fuk
ushi
ma/
2004
Swab
bH
AdV
-37
412.
3�
103
6.7
�10
98.
0�
108
HA
dV-3
799
.4H
AdV
-25
98.3
NT
AB
4372
76F
S161
/Fuk
ui/2
004
Isol
atea
HA
dV-3
72
NT
NT
NT
HA
dV-3
710
0.0
HA
dV25
98.9
NT
AB
3693
731/
Kyo
to/2
004
Swab
bH
AdV
-19a
62.
7�
103
3.3
�10
10
9.2
�10
9H
AdV
-25,
-26,
-38,
-49
97.4
HA
dV-2
4,-3
2,-4
297
.1N
TA
B43
7277
2/K
obe/
2004
Swab
bH
AdV
-37
23N
TN
TN
TH
AdV
-37
100.
0H
AdV
-25
98.9
NT
AB
4372
781/
Him
eji/2
004
Swab
bH
AdV
-37
7N
TN
TN
TH
AdV
-37
100.
0H
AdV
-25
98.9
NT
AB
4372
791/
Yam
aguc
hi/2
004
Swab
bH
AdV
-37
7N
TN
TN
TH
AdV
-37
100.
0H
AdV
-25
98.9
NT
AB
3693
722/
Miy
azak
i/200
4Sw
abb
HA
dV-3
79
NT
NT
NT
HA
dV-3
710
0.0
HA
dV-2
598
.9N
TA
B43
7280
aSp
orad
icin
fect
ion.
bN
osoc
omia
linf
ectio
n.c
NT
,not
test
ed.
3260 AOKI ET AL. J. CLIN. MICROBIOL.
knob, and penton genes derived from the HAdV-37, -22, -8,and -37 serotypes, respectively. We propose that this chimericvirus, HAdV-22,37/H8, be identified as a new causative agentof EKC.
MATERIALS AND METHODS
Patients and sample collection. We observed EKC outbreaks throughoutJapan between 2003 and 2004 (Weekly reports of adenovirus isolation/detectionfrom epidemic keratoconjunctivitis cases, 2001–2007; http://idsc.nih.go.jp/iasr/prompt/graph/ad3.gif) and collected a total of 171 conjunctival swabs from theEKC patients (Table 1). In 2003, 67 swabs were collected from sporadic infec-tions in five eye clinics from different parts of Japan: 12 swabs from Sapporo, 4from Tokyo, 22 from Matsuyama, 21 from Kumamoto, and 8 from Itoman. Nineswabs were obtained from patients with nosocomial infections in Tokyo. In 2004,93 swabs were obtained from six nosocomial outbreaks from different parts ofJapan: 41 swabs from Fukushima, 6 from Kyoto, 23 from Kobe, 7 from Himeji,7 from Yamaguchi, and 9 from Miyazaki. Two swabs were obtained from patientswith sporadic infections in Fukui. All swabs were positive for the AdV antigen bya lateral-flow immunochromatography assay (Adeno-check; Santen Pharmaceu-tical Co., Ltd., Osaka, Japan).
Viruses. One hundred microliters of the swab was added to A549 cells andincubated at 37°C. AdVs were identified by staining infected cells with an AdV-
specific monoclonal antibody (Chemicon International, Temecula, CA). Theisolates were further propagated in A549 cells without plaque purification forgenome typing and sequence analyses of full-length hexon, fiber, and pentonbase genes. The swabs and isolates from patients were supplied by the ReferenceCenter of Nosocomial Infection, Hokkaido University Graduate School of Med-icine, Sapporo, Japan. The prototype strains of 51 HAdVs were obtained fromthe American Type Culture Collection and the National Institute of InfectiousDiseases, Tokyo, Japan. These reference viruses were used directly for DNAextraction without further propagation.
Phylogeny-based classification using the partial hexon sequence. Viral DNAwas directly extracted from 100 �l of each conjunctival swab using a SumitestEX-R&D kit (Medical & Biological Laboratories Co., Ltd., Nagoya, Japan)according to the manufacturer’s instructions. After drying, the extracted DNAwas dissolved in 10 �l of TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA). A 350-bpsection of the hexon C4 nucleotide sequence was amplified as described previ-ously (22). In brief, PCR was carried out in a LightCycler Quick System 330(Roche) with a forward primer, AdnU-S�2 [nt 20743 to 20762; 5�-TTCCCCATGGC(A/T/C/G)CACAA(C/T)AC-3�], and a reverse primer, AdnU-A2 [nt 21274to 21296; 5�-TGCC(T/G)(A/G)CTCAT(A/G)GGCTG(A/G)AAGTT-3�]. Thepositions of the primers were numbered according to the complete nucleotidesequence of the HAdV-2 strain (GenBank accession no. J01917). The PCRprotocol was as follows: 95°C for 10 min for the initial activation of FastStart TaqDNA polymerase and the denaturation of template DNA, followed by 45 cycles
FIG. 1. Phylogenetic analyses of the representative strains C075/Matsuyama/2003, 8/Tokyo/2003, 1/Yamaguchi/2004, and FS-161/Fukui/2004(all marked with *) from patients with EKC. A 350-bp partial sequence of the hexon gene of the representative samples was analyzed by theneighbor-joining method together with those of the prototype strains of all 51 HAdV serotypes. The numbers at the nodes are percentages of 1,000bootstrap pseudoreplicates containing the cluster distal to the node.
VOL. 46, 2008 EKC DUE TO A HEXON-CHIMERIC-INTERMEDIATE HAdV 3261
of amplification, each consisting of denaturation at 95°C for 10 s, annealing at70°C for 10 s, and primer extension at 72°C for 25 s. The PCR products wereseparated on 3% agarose gel and purified with a QIAquick gel extraction kit(Qiagen, Valencia, CA). The nucleotide sequences of the hexon and fiber geneswere amplified as described elsewhere (21). PCR was performed in separaterooms for each PCR step (pre-PCR, specimen preparation and template addi-tion, and post-PCR) to prevent cross contamination. All samples were handledusing different pipettes with aerosol-resistant tips. Negative-control samples wereassayed in each PCR run (10). The number of copies of HAdV DNA in theclinical samples was calculated by a standard curve using pAd8hxn as the stan-dard, as described elsewhere (22).
The nucleotide sequences of the PCR products were determined using a CEQ
2000XL DNA analysis system with a DyeTerminator cycle sequencing kit (Beck-man Coulter, Fullerton, CA) and compared with those of all 51 serotypes usingSINCA (Fujitsu Limited, Tokyo, Japan). The evolutionary distances were esti-mated using Kimura’s two-parameter method (19), and unrooted phylogenetictrees were constructed using the neighbor-joining method (25). Bootstrap anal-yses were performed with 1,000 resamplings of the data sets (13). Similarity plotswere generated using SimPlot (version 3.5.1) (20), and the similarities werecalculated for each window of 200 nt by the Kimura two-parameter method (19),with a transition/transversion ratio of 2.0. The window was successively advancedalong the genome alignment in 20-nt increments.
Genome typing. Viral DNA was extracted from the infected cells in a 75-cm2
plastic flask using 3 ml of Hirt lysis solution (10 mM Tris, 1 mM EDTA, and 0.6%
FIG. 2. Genome type of the strain C075/Matsuyama/2003. The viral genomic DNA was digested with the restriction enzymes BamHI, BglI,BglII, HindIII, KpnI, SalI, XhoI, EcoRI, SacI, and SmaI. Lane 1, reference markers (EcoT14 and BglII digest of lambda DNA); lane 2, HAdV-37prototype strain; lane 3, C001/Matsuyama/2003, which was identified as HAdV-37; lane 4, C075/Matsuyama/2003.
3262 AOKI ET AL. J. CLIN. MICROBIOL.
sodium dodecyl sulfate at pH 8.0) (16). Proteinase K was added at a finalconcentration of 50 �g/ml, and the samples were incubated at 37°C for 1 h.Cellular DNA was precipitated with 1 M NaCl (final concentration) overnight at4°C. After phenol-chloroform extraction, the supernatant was treated with amixture of ribonucleases A (25 mg/ml) and T1 (80 U/ml) (Sigma, St. Louis, MO),and phenol-chloroform extraction was performed. Viral DNA was precipitatedwith isopropanol and suspended in 300 �l of TE buffer. One microgram of viralgenomic DNA was digested with 5 U of each of the restriction enzymes BamHI,BglI, BglII, HindIII, KpnI, SalI, XhoI, EcoRI, SacI, and SmaI (Takara ShuzoCo., Ltd., Kyoto, Japan). The digested viral DNA was loaded onto 1% agarosegels containing 1 �g/ml ethidium bromide. The DNA bands were photographedwith an UV transilluminator and a Polaroid camera. The migration patterns ofthe DNA fragments were compared with those of previously reported genometypes (1, 9, 15, 26, 30).
Serological analysis. Serological analyses were performed by a quantitativeNT with type-specific antisera (HAdV-8, -19, -37) purchased from Denka SeikenCo., Ltd. (Tokyo, Japan) or from the American Type Culture Collection (HAdV-8). The antiserum against HAdV-22 was kindly supplied from the Hiroshima CityInstitute of Public Health. The NT was performed with A549 cells in 96-wellmicroplates. The 100 50% tissue culture infective dose that caused a cytopathiceffect after 7 days of incubation at 37°C was used for the challenge virus.Duplicates of the twofold-serially-diluted antisera were used in the NT (27).
Nucleotide sequence accession numbers. The GenBank accession numbersobtained in this study are AB369371 to AB369373 and AB437260 to AB437280.GenBank sequences AB330082 to AB330132 were used to generate alignmentsof the hexon gene.
RESULTS
Phylogeny-based classification using a partial hexon C4 se-quence. We detected HAdV DNA from 171 conjunctivitis pa-tients whom we tested during 2003 to 2004. To assess thegenetic constellation, an alignment of the 350-bp partial hexonC4 nucleotide sequence of the HAdV DNA was performed forthe 51 prototype strains using the SINCA genetic softwareprogram. The 51 prototype strains showed 73.4 to 99.7% (av-erage, 87.1%) identity, with the exception of HAdV-11,HAdV-35, HAdV-21, and HAdV-50 (data not shown).
Out of 67 sporadic infections in 2003, 58 sequences (86.6%)were segregated into cluster D, 4 into cluster E, and 5 intocluster B (Fig. 1). Cluster D consists of 32 serotypes, HAdV-8to -10, -13, -15, -17, -19, -20, -22 to -30, -32, -33, -36 to -39, -42to -49, and -51, whose nucleotide identities range from 92.3%(between HAdV-8 and HAdV-28) to 99.7% (between HAdV-24
and HAdV-38, HAdV-25 and HAdV-38, and HAdV-32 andHAdV-38), with an average of 97.3%. Of the 58 cluster Dsequences, 41 (70.7%) presented 99.7 to 100% nucleotideidentity to the HAdV-37 prototype strain, 4 sequences pre-sented 100% nucleotide identity to HAdV-19a, and 13 se-quences presented 99.7 to 100% nucleotide identity to anotherrecently identified HAdV (17). All HAdV DNAs from noso-comial infections in Tokyo presented 100% nucleotide identityto HAdV-37. Eighty-seven of the 95 sequences (91.6%) fromthe patients in 2004 exhibited 99.4 to 100% nucleotide identityto HAdV-37, and 6 sequences exhibited identity to HAdV-19a(Fig. 1). Two HAdV DNAs from sporadic infections in Fukuishowed 100% nucleotide identity to HAdV-37. These resultsshowed that HAdV-37 was a major causative agent of EKC in2003 to 2004.
Genome typing. To determine the genotype of representa-tive HAdV isolates in cluster D from sporadic inflections in2003, the isolates were propagated in A549 cells (Table 1).Clinical isolates were inoculated into A549 cells in 75-cm2
plastic flasks. When the cytopathic effect was near completion,virus-infected cells were collected. Viral DNA was extractedfrom the infected cells as has been described elsewhere. Thegenomic DNA from the strain was digested with the 10 restric-tion endonucleases described in Materials and Methods. In-terestingly, one strain, C075/Matsuyama/2003, isolated in Mat-suyama City on Shikoku Island, showed a different restrictionpattern from that of the HAdV-37 prototype strain in cluster D(Fig. 2). The genome size of C075/Matsuyama/2003 calculatedby each fragment was approximately 3.5 kbp. These resultssuggested that the preparation did not contain a mixed viruspopulation.
Full-length hexon gene nucleotide sequences. To clarify thediscrepancy between the results of a phylogenetic analysisbased on the partial hexon C4 region and genome typing, weamplified the full-length hexon gene (nt 17784 to 20618) ofHAdV-37 using a pair of primers, AdVID (nt 17751 to 17772,5�-TGTATGTGCCTTACCGCCAGAG-3�) and AdL3D2 [nt20642 to 20661, 5�-GCGC(A/T)CGATGG(A/G)CGC(A/G)AGCT-3�], and determined the full-length hexon nucleotide
TABLE 2. Nucleotide sequence analysis of the L-1 and L-2 regions in the hexon gene and fiber knob regions
Region and virus (isolate/geographic origin/year of
isolation)
Highest-scoringprototype % Identity Next-highest-scoring
prototype % Identity% Identity to
HAdV-37prototype
% Identity toHAdV-22prototype
Hexon L-1 regionC075/Matsuyama/2003 HAdV-22 99.4 HAdV-49 70.8 61.08/Tokyo/2003 HAdV-22 99.4 HAdV-49 70.8 61.01/Yamaguchi/2004 HAdV-22 99.4 HAdV-49 70.8 61.0FS161/Fukui/2004 HAdV-22 99.4 HAdV-49 70.8 61.0
Hexon L-2 regionC075/Matsuyama/2003 HAdV-22 99.6 HAdV-9 82.8 77.08/Tokyo/2003 HAdV-22 100.0 HAdV-9 83.2 77.31/Yamaguchi/2004 HAdV-22 100.0 HAdV-9 83.2 77.3FS161/Fukui/2004 HAdV-22 99.6 HAdV-9 82.8 77.0
Fiber knob regionC075/Matsuyama/2003 HAdV-8 100.0 HAdV-9 94.5 83.9 59.28/Tokyo/2003 HAdV-8 100.0 HAdV-9 94.5 83.9 59.21/Yamaguchi/2004 HAdV-8 100.0 HAdV-9 94.5 83.9 59.2FS161/Fukui/2004 HAdV-8 100.0 HAdV-9 94.5 83.9 59.2
VOL. 46, 2008 EKC DUE TO A HEXON-CHIMERIC-INTERMEDIATE HAdV 3263
sequences. We also determined the full-length hexon nucleo-tide sequences of the 51 prototype strains. The nucleotideidentity of the 51 prototype strains ranged from 65.6% (be-tween HAdV-12 and HAdV-2) to 98.2% (between HAdV-15and HAdV-29), with an average of 78.9%. Nineteen sequencesfrom C075/Matsuyama/2003, two isolates (FS161/Fukui/2004and FS165/Fukui/2004) from sporadic infections, and nineswabs (8/Tokyo/2003) and seven swabs (1/Yamaguchi/2004)from nosocomial infections had full-length hexon gene nucle-otide sequences that were 99.4 to 100% identical. When full-length hexon genes of strains C075/Matsuyama/2003, FS161/Fukui/2004, FS165/Fukui/2004 (all from sporadic infections),8/Tokyo/2003, and 1/Yamaguchi/2004 (both from nosocomialinfections) were compared with those of the 51 serotypes,these five closely related strains showed only 89.6 to 89.7%nucleotide identity with HAdV-37, compared to 98.4% identitywith the HAdV-22 prototype strain (data not shown). Based onthe sequence analysis, the hexon gene was divided into fourconserved (C1 to C4) and three variable (V1 to V3) regions(11). The nucleic sequence of the 51 prototype strains varies inlength from 2,760 to 2,907 bp; however, the conserved regions
have a constant length (2,283 bp, total). The sequences ofC075/Matsuyama/2003, 8/Tokyo/2003, and 1/Yamaguchi/2004had a length of 2,808 bp, identical to that of HAdV-22 but notto that of HAdV-37 (data not shown).
Recently, the nucleotide sequences of the hexon L-1 and L-2regions, which contain the main NT ε determinant, were de-termined for all HAdV prototype strains, and criteria for typ-ing was proposed (21). Therefore we compared the nucleotidesequences of the L-1 and L-2 regions of C075/Matsuyama/2003, 8/Tokyo/2003, 1/Yamaguchi/2004, and FS161/Fukui/2004with those of the 51 prototype strains. The L-2 nucleotide se-quence of the isolates showed the highest identity to HAdV-22(99.6 to 100%) and only 77.0 to 77.3% nucleotide identity toHAdV-37 (Table 2). Therefore, the isolates formed a monophy-letic cluster with HAdV-22, not with the HAdV-37 prototypestrain (Fig. 3A). We also analyzed the L-1 region because of thehigher contribution to the ε determinant (21). The isolatesshowed 99.4% nucleotide identity to HAdV-22 but 61.0% identityto HAdV-37 (Table 2). The isolates formed a monophyletic clus-ter with the HAdV-22 prototype strain (Fig. 3B).
The inconsistent constellation of the loop region and the
FIG. 3. Phylogenetic analyses of the nucleic acid sequences of hexon L-2 (A), hexon L-1 (B), and the fiber knob (C). The nucleotide sequencesof the strains C075/Matsuyama/2003, 8/Tokyo/2003, 1/Yamaguchi/2004, and FS161/Fukui/2004 (all marked with *) together with those of the 51prototype strains of HAdV were analyzed by the neighbor-joining method. The numbers at the nodes are percentages of 1,000 bootstrappseudoreplicates containing the cluster distal to the node.
VOL. 46, 2008 EKC DUE TO A HEXON-CHIMERIC-INTERMEDIATE HAdV 3265
partial hexon C4 region suggested that recombination mighthave played an important role in the evolution of the isolates.To clarify the recombination events, we compared the fulllengths of the hexon genes of strains C075/Matsuyama/2003and 1/Yamaguchi/2004 with those of the 51 prototype strainsand further analyzed them by similarity plot analysis with asliding window of 200 residues (Fig. 4). When strain C075/Matsuyama/2003 was used as the query sequence and com-pared with the 51 prototype strains, its L-1 and L-2 regionswere very close to those of the HAdV-22 prototype strain,whereas the hexon C4 region was similar to that of theHAdV-37 prototype strain (Fig. 4A). 1/Yamaguchi/2004showed the same results as C075/Matsuyama (Fig. 4B). Theseresults indicate that the hexon gene of the isolates is a chimeraof HAdV-22 and -37.
Fiber and penton base nucleotide sequences. To date, sev-eral intermediate strains have been reported and are thought
to have emerged by recombination events in the hexon andfiber genes of strains of different serotypes (12, 23). Therefore,we determined the nucleotide sequences of the full-length fibergenes of the isolates (C075/Matsuyama/2003, FS161/Fukui/2004) and strains from swabs (8/Tokyo/2003, 1/Yamaguchi/2004, FS161/Fukui/2004). The nucleotide sequence showed99.9 to 100% identity with HAdV-8 but only 66.6 to 66.7%identity with HAdV-22 and 75.2 to 75.3% identity withHAdV-37 (data not shown). When the fiber knob nucleotidesequences of the isolates were compared with those of 51prototype strains, the isolates showed 100% identity withHAdV-8 but only 59.2% identity with HAdV-22 and 83.9%identity with HAdV-37 (Table 2). This phylogenetic analysisshowed the isolates forming a monophyletic cluster with theHAdV-8 prototype strain (Fig. 3C). The penton base nucleo-tide sequence of the isolates showed the highest identity(99.3%) with HAdV-37 (data not shown). Taken together,
FIG. 4. Similarity plots of the nucleotide sequences of the full-length hexon gene calculated by SimPlot 3.5.1. The full lengths of the hexon nucleotidesequences of C075/Matsuyama/2003 (A) and 1/Yamaguchi/2004 (B) were compared with those of the 51 prototype strains. Each point represents thesimilarity in nucleotide sequence between the query strain and the 51 prototype strains, within a sliding window of 200 nt, centered on the positions plottedand with a step of 20 residues between points. Positions containing gaps were excluded from analysis. The vertical axis indicates the nucleotide identitiesbetween the isolates and the 51 prototype strains, expressed as percentages. The horizontal axis indicates the nucleotide positions of the hexon gene. Thehorizontal axis at the bottom indicates the positions of conserved regions (C1 to C4), variable regions (V1 to V3), and L-1 and -2.
3266 AOKI ET AL. J. CLIN. MICROBIOL.
these results strongly support the idea that the isolates repre-sent a hexon-chimeric intermediate AdV, HAdV- 22,37/H8,likely to be a new causative agent of EKC.
Serological analysis. To examine the serological reactivitybetween the C075/Matsuyama/2003 and HAdV-D strains, aquantitative neutralization assay was performed with the anti-serum against HAdV-22 along with the antisera against theHAdV-8 and -37 prototype strains, which are causative agentsof conjunctivitis. C075/Matsuyama/2003 reacted with HAdV-22prototype-specific antiserum at a titer higher than 1:64 of the
homologous titer. Conversely, it did not react with antisera spe-cific for the HAdV-8 and -37 prototype strains. The low NT titerwas found in the reaction only with HAdV-8 antiserum (Table 3).Consequently, the new serotype has been defined on the basis ofits immunological distinctiveness, namely, a homologous neutral-ization titer/heterologous neutralization titer ratio of �16 in ei-ther direction (29). Therefore, these results suggest that the hexonL-1 and L-2 regions, which contain the main NT ε determinant ofC075/Matsuyama/2003, are from HAdV-22.
DISCUSSION
The classical typing of AdVs is performed by NT and HIusing type-specific antisera. In general, recombination eventsof different serotypes result in intermediate types with contra-dictory results by NT and HI. The molecular analysis ofHAdVs has shown that the ε determinant in the hexon loopregion is responsible for the neutralization properties, and the� determinant in the fiber knob region is responsible for hem-agglutination (14, 21, 24, 31). Therefore, to identify the inter-mediate strain by phylogenetic analysis, nucleotide sequence
FIG. 4—Continued.
TABLE 3. Quantitative NT results with C075/Matsuyama/2003against type-specific antisera
Virus strain
Quantitative neutralization titerof antisera for strain:
HAdV-22p HAdV-37p HAdV-8p
C075/Matsuyama/2003 4,096 �1 4HAdV-8p 8 1,024HAdV-22p 8,192HAdV-37p �1 128
VOL. 46, 2008 EKC DUE TO A HEXON-CHIMERIC-INTERMEDIATE HAdV 3267
analysis of both the hexon loop region and the fiber knobregion is required. An outbreak of EKC caused by a newintermediate AdV type 22/H8 has been reported in Germany(12). The hexon L-2 nucleotide sequence was identical to thatof HAdV-22, and the fiber knob nucleotide sequence was iden-tical to that of HAdV-8. This intermediate is very close to theone identified in this study.
In Japan, an intermediate HAdV has previously been iso-lated from patients with conjunctivitis and identified by NTand HI as a new causative agent of conjunctivitis, 22/H10,19,37(23). The strain was kindly supplied from the Hiroshima CityInstitute of Public Health. To clarify whether HAdV-22/H10,19,37 is a fiber-chimeric virus, we determined the nucle-otide sequences of the hexon L-2 region and the full-lengthsequence of the fiber genes and compared them with those ofHAdV-10, -19, and -37. The L-2 region was identical to that ofHAdV-22, whereas the nucleotide sequence of the full lengthof the fiber gene was identical to that of the HAdV-37 proto-type strain (data not shown). The intermediate HAdV-22/H10,19,37 strain might be cross-reactive by HI using type-specific antisera against HAdV-10 and -19 because thenucleotide sequences of the fiber knob regions of HAdV-10,-19, and -37 are phylogenetically related (Fig. 3C). Phylogeny-based identification using the hexon L-1 and L-2 regions dis-tinguished the 51 prototype strains by serotype. However, it didnot correctly classify the 51 prototype strains into six species;e.g., HAdV-4 (in HAdV-E) was classified into HAdV-B (Fig.3A and B), and HAdV-40 and -41 of HAdV-F were classifiedinto HAdV-A (Fig. 3B). We have recently determined thepartial hexon sequences of all 51 prototype strains and devel-oped a rapid and reliable method for diagnosis based on phy-logenetic analysis (22, 28). This method successfully classifiedthe 51 prototype strains of HAdVs into the six designatedspecies, as approved by the International Committee on Tax-onomy of Viruses; we have applied molecular diagnosis toidentify hundreds of isolates or swabs obtained over the last 30years from patients with EKC and lower respiratory tract in-fections from different parts of the world (5, 6, 18, 22, 28).Through such study, we found another novel HAdV fromnosocomially infected patients with EKC (17).
In Japan, we had four large outbreaks of EKC infectionsduring the period from 1990 to 2001 and identified HAdV-37as the causative agent by our methods (6). We also analyzedthe genome types of HAdV-37 isolates and found five newgenome types of HAdV-37. In 2003, we had an outbreak ofEKC due to HAdV-37 again. When we analyzed the genometypes of the representative isolates from each eye clinic, oneisolate had a restriction pattern different from that of theHAdV-37 prototype strain. To clarify the discrepancy of ourresults, we amplified full-length hexon and fiber genes andcompared their nucleotide sequences with those of the 51prototype strains. The present study is the result of this inves-tigation.
In conclusion, we here identified a new hexon-chimeric in-termediate AdV, HAdV-22,37/H8, which was associated withsporadic infection and nosocomial infection in 2003 to 2004.This is, to our knowledge, the first report of a hexon-chimericintermediate AdV causing an EKC outbreak. The origin andmeans of transmission of this strain are unknown. HAdV-22was isolated from a patient with trachoma in 1956 (7); how-
ever, it is rarely isolated from patients with EKC. In Japan, wehad an outbreak of EKC caused by HAdV-37 in 2003. It is wellknown that the ε determinant in the L-1 and L-2 regions isresponsible for the neutralization properties. The ability toescape immunity to HAdV-37 may have been acquired by arecombination event. HAdV-22,37/H8 should be monitored asa likely new causative agent of EKC; 3 years after the periodencompassed by this report (2003 to 2004), we additionallyfound an EKC nosocomial infection caused by this virus inSapporo, in the northern part of Japan, in 2007 (data notshown). We recommend determining the hexon L-1 or L-2region, partial C3 region, fiber knob, and penton base nucle-otide sequences to precisely identify isolates.
ACKNOWLEDGMENTS
We thank Noriko Inada (Division of Ophthalmology, Department ofVisual Science, Nihon University School of Medicine) and MasakoNakamura (Fukui Prefectural Institute of Public Health and Environ-mental Science) for providing the swabs and isolates from EKC pa-tients in Fukui prefecture.
REFERENCES
1. Adrian, T., G. Wadell, J. C. Hierholzer, and R. Wigand. 1986. DNA restric-tion analysis of adenovirus prototype 1 to 41. Arch. Virol. 91:277–290.
2. Aoki, K., M. Kato, H. Ohtsuka, K. Ishii, Nakazono, and H. Sawada. 1982.Clinical and aetiological study of adenoviral conjunctivitis, with special ref-erence to adenovirus type 4 and 19 infections. Br. J. Ophthalmol. 66:776–780.
3. Aoki, K., R. Kawana, I. Matsumoto, G. Wadell, and J. C. de Jong. 1986. Viralconjunctivitis with special reference to adenovirus type 37 and enterovirus 70infection. Jpn. J. Ophthalmol. 30:158–164.
4. Aoki, K., and Y. Tagawa. 2002. A twenty-one year surveillance of adenoviralconjunctivitis in Sapporo, Japan. Int. Ophthalmol. Clin. 42:49–54.
5. Ariga, T., Y. Shimada, K. Ohgami, Y. Tagawa, H. Ishiko, K. Aoki, and S.Ohno. 2004. New genome type of adenovirus serotype 4 caused nosocomialinfections associated with epidemic conjunctivitis in Japan. J. Clin. Micro-biol. 42:3644–3648.
6. Ariga, T., Y. Shimada, K. Shiratori, K. Ohgami, S. Yamazaki, Y. Tagawa, M.Kikuchi, Y. Miyakita, K. Fujita, H. Ishiko, K. Aoki, and S. Ohno. 2005. Fivenew genome types of adenovirus type 37 caused epidemic keratoconjuncti-vitis in Sapporo, Japan, for more than 10 years. J. Clin. Microbiol. 43:726–732.
7. Bell, S. D., Jr., T. Rondon Rota, and D. E. McComb. 1959. Adenovirusesisolated from Saudi Arabia. III. Six new serotypes. Am. J. Trop. Med. Hyg.8:523–526.
8. Benko, M., B. Harrach, G. W. Bothe, and W. C. Russel. 2005. FamilyAdenoviridae, p. 213–228. In C. M. Fauquet, M. A. Mayo, J. Maniloff, U.Desselberger, and L. A. Ball (ed.), Virus taxonomy. Eighth report of theInternational Committee on Taxonomy of Viruses. Elsevier Academic Press,San Diego, CA.
9. De Jong, J. C., A. G. Wermenbol, M. W. Verweij-Uijterwaal, K. W. Slaterus,P. Wertheim-Van Dillen, G. J. Van Doornum, S. H. Khoo, and J. C. Hier-holzer. 1999. Adenoviruses from human immunodeficiency virus-infectedindividuals, including two strains that represent new candidate serotypesAd50 and Ad51 of species B1 and D, respectively. J. Clin. Microbiol. 37:3940–3945.
10. Dragon, E. A., J. P. Sapadoro, and R. Madej. 1993. Quality control ofpolymerase chain reaction, p. 160–168. In D. H. Pering, T. F. Smith, F. C.Tenover, and T. J. White (ed.), Diagnostic molecular microbiology: princi-ples and applications. American Society for Microbiology, Washington, DC.
11. Ebner, K., W. Pinsker, and T. Lion. 2005. Comparative sequence analysis ofthe hexon gene in the entire spectrum of human adenovirus serotypes:phylogenetic, taxonomic, and clinical implications. J. Virol. 79:12635–12642.
12. Engelmann, I., I. Madisch, H. Pommer, and A. Heim. 2006. An outbreak ofepidemic keratoconjunctivitis caused by a new intermediate adenovirus22/H8 identified by molecular typing. Clin. Infect. Dis. 43:e64–66.
13. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach usingthe bootstrap. Evolution 39:783–791.
14. Gall, J. G., R. G. Crystal, and E. Falck-Pedersen. 1998. Construction andcharacterization of hexon-chimeric adenoviruses: specification of adenovirusserotype. J. Virol. 72:10260–10264.
15. Hierholzer, J. C., R. Wigand, L. J. Anderson, T. Adrian, and J. W. M. Gold.1988. Adenovirus from patient with AIDS: a plethora of serotypes and adescription of five new serotypes of subgenus D (types 43–47). J. Infect. Dis.154:804–813.
3268 AOKI ET AL. J. CLIN. MICROBIOL.
16. Hirt, B. 1967. Selective extraction of polyoma DNA. J. Mol. Biol. 26:365–369.
17. Ishiko, H., Y. Shimada, T. Konno, A. Hayashi, T. Ohguchi, Y. Tagawa, K.Aoki, S. Ohno, and S. Yamazaki. 2008. Novel human adenovirus causingnosocomial epidemic keratoconjunctivitis. J. Clin. Microbiol. 46:2002–2008.
18. Jin, X. H., H. Ishiko, T. H. Nguyen, T. Ohguchi, M. Akanuma, K. Aoki, andS. Ohno. 2006. Molecular epidemiology of adenoviral conjunctivitis inHanoi, Vietnam. Am. J. Ophthalmol. 142:1064–1066.
19. Kimura, M. 1980. A simple method for estimating evolutionary rates of basesubstitutions through comparative studies of nucleotide sequence. J. Mol.Evol. 16:111–120.
20. Lole, K. S., R. C. Bollinger, R. S. Paranjape, D. Gadkari, S. S. Kulkarni,N. G. Novak, R. Ingersoll, H. W. Sheppard, and S. C. Ray. 1999. Full-lengthhuman immunodeficiency virus type 1 genomes from subtype C-infectedseroconverters in India, with evidence of intersubtype recombination. J. Vi-rol. 73:152–160.
21. Madisch, I., G. Harste, H. Pommer, and A. Heim. 2005. Phylogenetic anal-ysis of the main neutralization and hemagglutination determinants of allhuman adenovirus prototypes as a basis for molecular classification andtaxonomy. J. Virol. 79:15265–15276.
22. Miura-Ochiai, R., Y. Shimada, T. Konno, S. Yamazaki, K. Aoki, S. Ohno, E.Suzuki, and H. Ishiko. 2007. Quantitative detection and rapid identificationof human adenoviruses. J. Clin. Microbiol. 45:958–967.
23. Noda, M., Y. Miyamoto, Y. Ikeda, T. Matsuishi, and T. Ogino. 1991. Inter-mediate human adenovirus type 22/H10,19,37 as a new etiological agent ofconjunctivitis. J. Clin. Microbiol. 29:1286–1289.
24. Pring-Akerblom, P., and T. Adrian. 1994. Type- and group-specific polymer-ase chain reaction for adenovirus detection. Res. Virol. 145:25–35.
25. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new methodfor reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425.
26. Schnurr, D., and M. E. Dondero. 1993. Two new candidate adenovirusserotypes. Intervirology 36:79–83.
27. Schrader, E., and R. Wigand. 1981. Neutralization of adenovirus infectivityand cytotoxin in various cell cultures. J. Virol. Methods 2:321–330.
28. Shimada, Y., T. Ariga, Y. Tagawa, K. Aoki, S. Ohno, and H. Ishiko. 2004.Molecular diagnosis of human adenoviruses d and e by a phylogeny-basedclassification method using a partial hexon sequence. J. Clin. Microbiol.42:1577–1584.
29. Wigand, R., A. Bartha, R. S. Dreizin, H. Esche, H. S. Ginsberg, M. Green,J. C. Hierholzer, S. S. Kalter, J. B. McFerran, U. Pettersson, W. C. Russell,and G. Wadell. 1982. Adenoviridae: second report. Intervirology 18:169–176.
30. Wigand, R., T. H. Adrian, and F. Bricout. 1987. A new human adenovirus ofsubgenus D: candidate adenovirus type 42. Arch. Virol. 94:283–286.
31. World, W. S. M., and M. S. Horowitz. 2007. Adenoviridae, p. 2395–2436. InD. M. Knipe et al. (ed.), Fields virology, 5th ed., vol. II. Lippincott Williams& Wilkins, Philadelphia, PA.
VOL. 46, 2008 EKC DUE TO A HEXON-CHIMERIC-INTERMEDIATE HAdV 3269