A multigene typing system for human adenoviruses reveals a new genotype in a collection of Swedish clinical isolates
A multigene typing system for human adenoviruses reveals a new genotype in a collection of Swedish clinical isolates
Gy?z? L a?szl o? Kaj a?nID 0 1
Agnieszka Lipiec 0 1
D a?niel Bartha 1
Annika AllardID 0 1
Niklas Arnberg 0 1
0 Department of Clinical Microbiology, Virology, and Laboratory for Molecular Infection Medicine Sweden (MIMS), Umea? University, Umea?, Sweden, 2 Institute for Veterinary Medical Research, Centre for Agricultural Research, Hungarian Academy of Sciences , Budapest , Hungary
1 Editor: Ilya Ulasov, Sechenov First Medical University , RUSSIAN FEDERATION
Human adenoviruses (HAdVs) are common pathogens that can cause respiratory, gastrointestinal, urogenital, and ocular infections. They are divided into seven species containing 85 genotypes. Straightforward typing systems might help epidemiological investigations. As homologous recombination frequently shapes the evolution of HAdVs, information on a single gene is seldom sufficient to allow accurate and precise typing, and complete genomebased methods are recommended. Even so, complete genome analyses are not always easy to perform for practical reasons, and in such cases a multigene system can provide considerably more information about the strain under investigation than single-gene-based methods. Here we present a rapid, generic, multigene typing system for HAdVs based on three main deterministic regions of these viruses. Three PCR systems were used to amplify the genes encoding the DNA polymerase, the penton base hypervariable Arg-Gly-Asp-containing loop, and the hexon loop 1 (hypervariable region 1-6). Using this system, we typed 281 clinical isolates, detected members of six out of seven HAdV species (Human mastadenovirus A-F), and could also detect not only divergent strains of established types but also a new recombinant strain with a previously unpublished combination of adenovirus genomes. This strain was accepted by the Human Adenovirus Working Group as a novel genotype: HAdV-86. Seven strains that could not be typed with sufficient accuracy were also investigated using a PCR based on part of the fiber gene. By analysis of corresponding sequences of the 86 known HAdV genotypes, we determined that the proposed typing system should be able to distinguish all non-recombinant types, and with additional fiber information, all known HAdV genotypes.
Funding: This project was made possible by
funding from FP7 Marie Curie Actions via the
ADVEC consortium (participant: NA, grant
agreement no.: 324325). The funder had no role in
study design, data collection and analysis, decision
to publish, or preparation of the manuscript.
The seven species of human adenovirus (Human mastadenovirus A?G; HAdV-A to HAdV-G)
are divided into types, originally called serotypes. Virus strains were traditionally classified
into serotypes on the basis of serum neutralization tests [
]. These approaches are rather
tedious and time consuming, and require numerous reference strains and antisera. Because of
this, the classification of adenoviruses currently relies on molecular typing methods [
New or rare HAdV types may play a crucial role in industrial applications, as many known/
common human adenovirus types are already patented, or pre-existing immunity limits their
use in vector development. The most common HAdV-5 based vectors have been associated
with liver- and innate immunity-associated toxicity [
]. Even so, adenovirus-based vectors
have been developed for the treatment of both cancer and cardiovascular diseases, and for
prevention of infectious diseases. Vector engineering is a very active and evolving field in virology
To find possible new vector candidates (e.g. new and/or rare human adenovirus types), we
have used molecular methods to type 281 human adenovirus strains that were isolated from
patients in Sweden between 1978 and 2010. To ensure that we would not miss discovering any
new recombinant genotypes, we used a multigene approach.
The external surfaces of the viral major capsid proteins?the hexon, the penton base, and
the fiber [
]?contribute to antigenicity [
]; but the hexon loop 1 function as the major
antigenic determinant that is mostly responsible for serotyping results . The genes of these
three proteins are therefore used in the description and characterization of new recombinant
HAdV strains [
], so these would also be the best choice for multiple gene-based typing of
HAdV strains. Apart from using the hexon gene and the penton base gene, a fiber-targeted
PCR system was also tested using seven clinical isolates [
]. With six primer pairs, this
multiplex PCR had been developed for species HAdV-A?F, but unfortunately it gave a low positivity
rate (50%), and was therefore discarded from the set of PCR systems used. Also, since
information from the DNA polymerase gene is required to classify adenoviruses into species [
decided to use the hexon gene, the penton base gene, and the polymerase gene.
Materials and methods
Origin of strains
The 281 HAdV clinical isolates analyzed were from Sweden: from Sk?ne University Hospital,
Lund (n = 126) and from Norrland University Hospital, Ume? (n = 155). The strains had
originated from diverse clinical conditions, and had been isolated on A549, Vero, Madin-Darby
canine kidney, green monkey kidney or MA-104 cells. The original isolates were investigated
without any further propagation.
DNA was extracted from 200 ?L of isolate using a Qiagen BioRobot M48 workstation and the
Qiagen MagAttract DNA Mini M48 Kit according to the manufacturer?s protocol.
PCR systems and sequencing
Three PCR systems were used to type the adenovirus samples, targeting the genes of the viral
DNA polymerase, the hypervariable sequence encoding the Arg-Gly-Asp-containing loop of
the penton base, and a part of the hexon gene that encodes loop 1 (hypervariable region 1?6).
The latter nested PCR was based on the work of Lu and Erdman [
]. The primers of the first
two systems were designed based on publicly available conserved consensus sequences of
HAdV-A to HAdV-G. The primer sequences are summarized in Table 1. The exact
constitution and thermal profiles of the PCR reactions are detailed in S1 and S2 Tables.
2 / 15
Corresponding base pair ordinals of the HAdV-5
Without gel electrophoresis, the PCR products were sent to a commercial sequencing
provider (Macrogen Europe) for Sanger sequencing. The sequences of the products were
determined on both strands. As visualization of the PCR products was omitted, a PCR reaction was
considered to be positive if the resulting assembled sequence contig measured 508?511 bp for
the DNA polymerase-targeted PCR, 293?494 bp for the penton base-targeted PCR, and 714?
855 bp for the hexon-targeted PCR. Otherwise, the reaction was considered to be negative.
Mixed isolates were considered to be positive, but they were excluded from phylogenetic
Base calling and assembly of DNA polymerase and hexon sequences was performed using
], and the assembly of penton base sequences was performed using Cap3 [
Recombinant genotypes are described using their penton base, hexon, and fiber sequences,
and correct typing was difficult in some cases without having fiber-based data. Thus, we also
performed a fiber gene-targeted PCR [
] for seven of the strains where the other three regions
did not allow us to type the strain unambiguously (UmU053, UmU061, UmU082, UmU219,
UmU234, UmU262, and UmU271).
Rough typing of strains based on partial sequences
The strains were typed molecularly based on the derived amino acid sequence of all three (or
four) sequence stretches. For all stretches amplified, the most closely related (i.e. with highest
sequence identity, in other words, the least distant) HAdV reference strain was determined
using the pairwise-alignment-based Sequence Demarcation Tool (SDT) v1.2 [
Each reference strain (also called prototype) represented a different HAdV type. Up to
serotype ordinal 41 (HAdV-41), the strains recommended by Horwitz [
] were used. For types
HAdV-53 to HAdV-84, we used the strains recommended by the Human Adenovirus
Working Group (http://hadvwg.gmu.edu/). For types HAdV-42 to HAdV-52, we chose a strain with
the complete genome sequence available.
As DNA polymerase and penton base sequences were not sufficiently variable (i.e. more
reference strains shared identical or very closely related sequence stretches), a type assessment
was reached, in most cases based on the hexon result. If this raised the possibility that the strain
represented a previously described recombinant type, the sequence identity values of the DNA
polymerase and penton base were also evaluated carefully (see example below). In such a case,
we investigated whether these differed from the hexon result, which would suggest a final type
assessment as a recombinant genotype. For example, the hexon amplimer of strain UmU044
showed the highest sequence identity with that of HAdV-64 (S3 Table), and the second highest
with that of HAdV-19 (98.5%). Its DNA polymerase amplimer was identical to those of
HAdV-37, -60, and -64, and its penton base amplimer was identical to those of HAdV-22, -42,
3 / 15
-59, and -64. As HAdV-64 is a recombinant of HAdV-22, -19, and -37 (penton base, hexon,
and fiber, respectively) [
], UmU044 was typed as HAdV-64.
To be able to identify divergent HAdV strains, we determined the minimum level of
divergence between two distinct, known serotypes. That is, the two most closely related
non-recombinant HAdV reference strains (serotypes) were determined based on all three sequence
stretches corresponding to the PCR products. Primers were omitted from the analysis. The
corresponding base pair ordinals of the HAdV-5 genome (AC_000008) were 5217?5686 for
the DNA polymerase, 14917?15361 for the penton base, and 19192?19958 for the hexon
Complete genome sequences
Based on the typing results, four divergent strains (UmU010, UmU193, UmU225, and
UmU253) were identified. Furthermore, UmU018 had a potentially novel genome
composition that had not been seen previously. For further characterization, these five clinical isolates
were propagated on human alveolar epithelial cells (A549). Intracellular viral DNA was
purified from infected cells using the protocol of Kajon and Erdman [
]. The genomes of
UmU010, UmU193, and UmU225 were sequenced using Ion Torrent next-generation
sequencing at the Uppsala Genome Center of the National Genomics Infrastructure
(SciLifeLab; Uppsala, Sweden). We did not succeed in producing the required amount of DNA of
UmU018 and UmU253 in the first round, and for the quality required, UmU193 also required
supplementation with additional sequence information, so these three strains were sequenced
in the next round using Illumina HiSeq2500 technology at GATC Biotech (Konstanz,
Germany). The resulting reads were normalized to a 60-times coverage using BBNorm from the
BBTools suite. The normalized reads were assembled de novo using Mira version 4.9.5_2 [
and the original sequence reads were mapped to the resulting consensus sequences using the
Geneious mapper at the highest sensitivity with five iterations in Geneious 9.1.8 [
]. The new
consensus sequences were annotated based on HAdV reference strain genome annotations,
using the Annotate & Predict function of Geneious. The annotations were checked manually
The sequences from the virus strains were deposited in GenBank (NCBI; accession numbers:
Phylogenetic analysis of complete genomes
To investigate their phylogenetic relationship, the completely sequenced strains were analyzed
using both multiple-alignment-based phylogenetic tree inference and
pairwise-alignmentbased sequence identity calculation. Both types of analyses were based on complete genome
sequences and derived amino acid sequences of the entire hexon, penton base, and DNA
polymerase; and also hexon loop 1 (delimited according to Yuan et al. [
]) and the fiber knob. As
UmU253 had two fibers, both fiber knobs were analyzed.
For phylogenetic tree inference, multiple alignments were conducted using MAFFT [
and phylogenetic calculations were performed using RAxML 8.2.10 [
] based on alignments
edited in Gblocks 0.91b [
]. Evolutionary model selection for the complete genome sequence
alignment was performed using MEGA 7 [
], and using RAxML for the protein alignments.
The robustness of the trees was determined with a non-parametric bootstrap calculation using
1,024 repeats. To accelerate phylogenetic analyses, the calculations were run in parallel on 32
processor cores. RAxML always calculates integer multiples of the cores used, and uses the
4 / 15
smallest value above the desired one as the number of replicates. In this case, 1,000 replicates
were targeted, so 1,024 (32 32) were used. Phylogenetic trees were visualized using MEGA 7
], trees were rooted on the midpoint, and bootstrap values are given as percentages if they
Pairwise-alignment-based sequence identity calculations were conducted using the MAFFT
alignment algorithm in SDT [
]. The most closely related HAdV type (reference strain)
was determined for each completely sequenced strain, based on all sequence stretches
analyzed. Next, those non-recombinant serotypes were indicated which showed the highest level
of sequence identity with each other. It is important to stress the word ?serotype? here. By
determining the minimum divergence between two distinct non-recombinant serotypes, we
could evaluate the level of divergence of the completely sequenced strains.
Recombination events in strain UmU018 were analyzed further using SimPlot 3.5.1 [
based on the complete genome alignment and subslices corresponding to the penton base
gene, the hexon gene, and the E3 region.
Efficiency of PCR amplification
Using the PCR systems targeting the hexon gene, the polymerase gene, and the penton base
gene, we found that the virus strains gave 99.3%, 94.7%, and 77.9% positivity, respectively, and
that 5.4%, 6.4%, and 2.3% of the sequences analyzed were mixtures. Such mixture reads were
considered to be positive but we excluded them from the subsequent phylogenetic analysis.
Rough typing results based on partial sequences
The results of typing of the strains are summarized in S3 Table. Analysis of part of the DNA
polymerase gene revealed that 71.5% of the strains that gave DNA polymerase sequences
belonged to species HAdV-C, but almost all other human adenovirus species were found as
well, with the exception of species G. In the samples, 18 different HAdV types were identified;
the most common being HAdV-2 (38.1%) and HAdV-1 (22.8%). We conducted preliminary
analyses comparing the calculated sequence identities (S3 Table) to the ones measured
between different HAdV types (data not shown). These indicated that four strains (UmU010,
UmU193, UmU225, and UmU253) diverged significantly from the closest HAdV reference
strain, so the four complete genomes were sequenced.
At least two of the three PCR analyses provided sequence information for 251 strains, and
seven of these were found to be recombinants. Furthermore, as UmU114 was typed as
HAdV60 based on the hexon gene only, this strain was also classified as a recombinant (S3 Table).
Most of these recombinants were already accepted HAdV types, e.g. HAdV-64 (n = 3).
In the beginning of this study, recombinant strains were identified erroneously, based on
the DNA polymerase, penton base, and hexon sequences only. For example, UmU053 was
identified first as HAdV-66 (a recombinant genotype), as its penton base showed highest
amino acid identity to HAdV-7, and its hexon to HAdV-66. But the genomic composition of
HAdV-66 is an HAdV-7 penton base, an HAdV-7 hexon, and an HAdV-3 fiber. So, to check
this result, the fiber gene sequence of UmU053 (and also six other strains: UmU061, UmU082,
UmU219, UmU234, UmU262, and UmU271) was also determined, which contradicted the
previous typing result (in the other cases too). These strains were originally typed as
recombinant HAdV-66 or HAdV-68 based on their hexon sequences, but the HAdV-3 or HAdV-7
fiber sequences showed that these strains were common HAdV-3 (UmU219, UmU234,
UmU262, and UmU271) or HAdV-7 (UmU053, UmU061, and UmU082). Without fiber data,
5 / 15
such strains could not be typed accurately, so two types are given as the final type identified
(S3 Table), e.g. HAdV-7 or HAdV-66 for UmU062.
UmU018 had a new combination of adenovirus genomes. The hexon loop 1 of this strain
showed high identity (99.2%) to HAdV-25, its penton base showed high identity to HAdV-9,
-10, -26, and -56 (99.0%), and its DNA polymerase showed high identity to 37 different
HAdV-D types, including HAdV-9, -10, -25, and -26 (98.7?99.4%). This strain was chosen for
complete genome sequencing.
Complete genomes and their phylogenetic analysis
The read-coverage reports and basic attributes of the genomes are summarized in S4 Table.
All phylogenetic tree inferences (Fig 1) and sequence identity analyses (Table 2) confirmed
that UmU010, UmU193, UmU225, and UmU253 were most closely related to HAdV-12,
HAdV-5, HAdV-4, and HAdV-41, respectively. The only two exceptions were the DNA
polymerase and penton base analyses of UmU193, where HAdV-6 was found to be most closely
related to this strain, though HAdV-5 also gave very high degrees (percentages) of identity.
The recombinant strain UmU018 had a DNA polymerase that was most similar to that of
HAdV-48, -58, and -65, whereas its penton base showed high amino acid sequence identity to
HAdV-9, -10, -56, and -82. To find the most closely related HAdV type, the stretch of penton
Fig 1. Phylogenetic analysis of the five completely sequenced human adenovirus strains. The complete genome analysis was based on nucleotide sequences; all other
analyses were based on derived amino acid sequences.
6 / 15
The HAdV serotype (reference strain) most closely related to the completely sequenced human adenovirus strains was determined. This comparison was conducted on
six sequence stretches: complete genome nucleic acid (row 2), derived amino acid sequences of the complete DNA polymerase, penton base, and hexon (rows 3?5), and
also the derived amino acid sequences of hexon loop 1, and the fiber knob (rows 6?7). As strain UmU253 had two fibers, both fiber knobs were analyzed. In the two
cases where UmU193 showed highest sequence identity to HAdV-6, the HAdV-5 percentages are also indicated. The most closely related non-recombinant HAdV
serotypes (reference strains) were also determined based on all sequence stretches analyzed: here meaning the serotypes most closely related to each other (coloumn 2).
Human adenovirus types are represented by their type number only. Abbreviations: AA, amino acid; comp., compared; compl., complete; F1, fiber 1; F2, fiber 2; HAdV,
human adenovirus; NA, nucleic acid.
The same sequence identity was measured compared to the derived DNA polymerase amino acid sequence of HAdV-48, -58, and -65.
? The same sequence identity was measured compared to the derived penton base amino acid sequence of HAdV-9, -10, -56, and -82.
base was also analyzed at the nucleic acid level, and showed the highest level of sequence
identity to HAdV-9 (99.17%). The hexon and fiber proteins of UmU018 were most similar to those
of HAdV-25 (Figs 1 and 2, and Table 2). Furthermore, SimPlot analysis revealed that its E3
genomic region was a recombinant of HAdV-26, -51, and -82 (Fig 2); and the derived amino
acid sequences of the DNA polymerase, core protein V, and the 100 K protein showed high
levels of sequence identity to several species D types (Table 3).
The five sequenced strains were evaluated by the Human Adenovirus Working Group as
possible new HAdV type candidates, and UmU018 (complete strain designation: Adenovirus
D human/SWE/UmU018/1978/86[P9H25F25]) was approved as HAdV genotype 86.
In this era of quick, high-throughput, and cheap sequencing methods, one might easily
question the rationale behind using a PCR-based typing system for HAdVs, especially when we
consider the very limited amount of information gained by using it. Even so, at least at the
present time, no real practical alternatives are available for the screening of hundreds of virus
strains. Target enrichment and genome sequencing methods do improve, become cheaper,
and certainly provide abundant information about the strain under examination: this is
undoubtedly the future of virus typing. However, available target enrichment methods require
the PCR amplification of the enriched fragment pool [
], so they do not provide any real
advantages over the method detailed here. We also considered propagating all strains to obtain
the amounts of DNA required for complete genome sequencing. However, this would have
been far too expensive and time consuming. Thus, we used this PCR-based screening method
A quick HAdV typing system would benefit researchers, e.g. epidemiologists, as at least a
partial genomic composition of an HAdV strain could be determined. Although there are
several typing systems available, many HAdV typing systems are limited: (1) some systems are
7 / 15
Fig 2. Genomic layout and recombination analysis of human adenovirus type 86 (HAdV-86), strain UmU018. Green arrows in the genome map represent protein
coding sequences, red arrows represent virus-associated RNAs, and brown arrows represent the inverted terminal repeats. Human adenovirus types showing the highest
sequence identity in characteristic domains or coding sequences are shown abbreviated (e.g. HAdV-9 means human adenovirus 9).
not generic [
]; (2) diverse HAdV types may be detected using multiple PCRs [
3, 4, 20, 39,
]; (3) the antigenic determinant loops may be excluded [
5, 20, 41
]; (4) even if the loops are
included, the product may be too long for convenient sequence analysis [
recombinant strains can only be detected by a very complex algorithm [
]; or (6) the system may
target only one gene, excluding the possibility of recombinant detection [
5, 21, 38, 40?42, 46
Our multigene HAdV typing system provides a solution to the six issues mentioned above by
providing sequence information from three characteristic regions of the HAdV genome.
Nowadays, if information about more than a single gene is needed, this is usually achieved
by sequencing of the complete genome [
]. Although this approach provides abundant
information about the strain being examined, it is still more time consuming (requiring virus
isolation) and more expensive than a PCR-based multigene typing system. Isolates were
analyzed in our study also, but DNA derived from clinical material is often used in many
PCRbased diagnostic methods. The hexon gene-targeted PCR, for example, was tested in the
original publication and also later using clinical specimens [
]. Thus, it is likely that our typing
system would give at least partial positivity using clinical material, and this is its main
advantage over contemporary complete genome typing methods, as virus isolation and propagation
can be time consuming.
48, 58, 65
17, 24, 32, P38H32F27
10, 44, 56, 72, P67H9F15
8 / 15
As the hexon gene-targeted PCR had been validated previously using a panel of 51 HAdV
], we did not conduct such an analysis. In the present study, the penton base-targeted
PCR had the lowest positivity rate (77.9%), but this result might only have been a consequence
of the limited PCR sensitivity, possibly caused by highly degenerate primers. The possibility of
selective specificity was excluded, as every HAdV species was detectable using it. Using the
typing system, we detected 6 HAdV species among the samples; HAdV-G was the exception. The
hexon primers had been designed based on HAdV-A to HAdV-F sequences only [
However, as the penton base gene- and DNA polymerase gene-targeted PCRs were designed based
on reference sequences from HAdV-A to HAdV-G, it is quite probable that all HAdV-G
strains would yield a positive result using these. Thus, our generic HAdV multigene typing
system is capable of providing sequence information from all HAdV species on (1) the hexon
loop 1, (2) the hypervariable loop of the penton base, and (3) the DNA polymerase?the first
two of which are antigenic determinants [
] and the latter is the main species determinant
Besides sensitivity, another important factor for typing systems is their resolution capacity.
Being the major antigenic determinants, hexon loops 1 and 2 are often used in molecular
typing of HAdV strains [
3, 4, 42, 46, 50
] and provide a type-specific sequence for the
non-recombinant HAdV serotypes. This was confirmed in the study providing the hexon gene-targeted
primers, where 51 reference adenovirus strains (formerly demarcated using serotyping) were
]. The same homologous sequence stretch, originating from the same and newer
reference strains, was analyzed in silico in our study (now representing 86 HAdV genotypes),
providing the same result: non-recombinant HAdV strains can be typed based on hexon loop
1 amino acid sequences; that is, non-recombinant HAdV types can be determined using this
sequence stretch. Combining the hexon data with penton base and fiber results, we could
achieve a preliminary typing of any HAdV strain, even recombinant ones, as all accepted
HAdV genotypes have a unique combination of penton base, hexon, and fiber genes [
Unfortunately, fiber data could not be obtained with our approach, and some closely related
types share identical or very similar penton base stretches, as described already by Ismail et al.
]. This makes distinction impossible for the following ten pairs of 19 genotypes using only
DNA polymerase, penton base, and hexon information: HAdV-3 and -68, -7 and -66, -11 and
-55, -11 and -78, -15 and -29, -20 and -60, -21 and -76, -30 and -63, -56 and -82, and finally -77
and -79. All other recombinant or recently described HAdV types (HAdV-16, -53, -54, -57?59,
-61, -62, -64, -65, -67, -69?75, -80, -81, and -83?86) can be distinguished from the parental
types and any other types using the combination of hexon and penton base amino acid
In conclusion, this typing system is not entirely optimal to completely describe or type an
HAdV strain. For definite typing, complete genome information is required [
homologous recombination is one of the major driving forces in HAdV evolution [
Alternatively, if not the complete genome, at least fiber information is crucial in addition to
the hexon and penton base sequences [
]. However, the system developed is adequate for
preliminary typing in some cases. In addition, a new recombinant HAdV strain was found using
this typing system, as discussed in detail below; furthermore, divergent strains of established
types were also detected in various HAdV species. Users must weigh the advantages and
disadvantages of the methods available, as well as the time available, the budget, and the number of
Of the strains analyzed, we observed a very high proportion of HAdV-C strains, but almost
all other human adenovirus species (HAdV-A to HAdV-F) were also found. A similar
predominance of HAdV-C strains was observed by Sriwanna et al. [
]. Al Qurashi et al. also
detected strains belonging to six HAdV species, and a similar high proportion of HAdV-C, but
9 / 15
there was a higher proportion of species HAdV-A types as compared to our results [
other studies, more diverse strain distribution or a higher proportion of HAdV-B was observed
21, 58, 59
]. The diverse nature of the clinical samples investigated in this study might explain
this result: our study was not limited to respiratory samples, for example.
Mixture sequences originate from co-infections, which happens often [
]. Based on hexon
sequences, we found that 5.4% of the isolates that gave a positive PCR reaction contained at
least two different HAdV types, which is similar to the 3.5% observed by McCarthy et al. [
but lower than the 18.2% observed by Metzgar et al. [
]. In eight cases out of the 15, the DNA
polymerase sequence could be interpreted, showing that the co-infecting strains belonged to
the same species.
Every strain that was typed as HAdV-5 based on the hexon loop 1 sequence had a penton
base stretch closest to that of HAdV-1, and 90 of the 99 strains typed as HAdV-2 based on the
hexon loop 1 sequence also had a DNA polymerase stretch closest to that of HAdV-1. As there
is a mere two-amino-acid difference between the penton base loops of HAdV-1 and HAdV-5
and a one-amino-acid difference between the HAdV-1 and HAdV-2 DNA polymerase stretch
analyzed, these strains were not regarded as recombinants. As previously observed in penton
base sequences [
], DNA polymerase sequences might also be enough only for species
In the strains examined, we also tried to identify possible new HAdV candidate types. We
wanted to enable the discovery of possible new types based on both traditional and
contemporary techniques, and this could be achieved by our multigene molecular typing system. We
aimed to identify (1) new, divergent, distance-based types, showing low sequence identity to
the closest related reference strain?earlier this would have been a new serotype, and (2) new
and unpublished recombinants?new genotypes.
Combining the three different typing results, we discovered one possible new recombinant,
UmU018, with a previously unpublished combination of HAdV genomes. This strain has been
approved by the Human Adenovirus Working Group to represent a novel genotype:
HAdV86. Similar to previous findings in the case of HAdV-84 [
], HAdV-86 has a general HAdV-D
backbone: conserved core proteins show very high sequence identity to several different closely
related HAdV-D types. The surface proteins with antigenic determinants showed a close
relationship to those of HAdV-9 and -25. Perhaps the exact direction of evolution cannot be
determined based on these similarities, e.g. the penton base loops may have originated from
HAdV9, or HAdV-9 may have inherited these from HAdV-86, or both of these types inherited them
from a common, but unknown ancestor. Homologous recombination has a crucial role in
HAdV evolution [
], as it provides the progeny strain with a new set of major antigenic
determinants. This enables a quick antigenic shift in a possibly neutralizing environment, or
the use of novel receptors. The E3 region is also a hotspot for recombinations in the HAdV
genome . Most of the HAdV-86 strain?s E3 region is closely related to that of HAdV-26,
but the CR1-gamma represents a novel proteotype. The amino acid sequence of the
CR1-gamma protein shows only 77.9?81.5% identity to that of HAdV-22, -37, or -82, and the
threshold for a novel proteotype was set at 10% divergence [
]. UmU018 was isolated in 1978;
the only information available suggests that the patient was investigated for infection by
The other four sequenced strains cannot be recognized as new HAdV types, as the imputed
serology used does not always correlate strongly with serotyping, and the minor sequence
differences observed might not represent sufficient differences in a conventional serum
neutralization assay. Also, recognizing these strains as new types could cause considerable confusion
in the field. As a member of the Human Adenovirus Working Group pointed out, for example,
UmU225 is almost identical to field strains NHRC 3 and NHRC 42606 (complete genome
10 / 15
nucleic acid identity: 99.85% and 99.96%, respectively), both of which were typed as HAdV-4,
and historically within this serotype as genome type HAdV-4a [
]. A virtual restriction
endonuclease cleavage analysis of the UmU225 genome also supported classification of the strain as
HAdV-4a (data not shown). Although this genome type is divergent from the reference strain
(RI-67, genome type: HAdV-4p, ?p? meaning prototype or reference strain), it can still be
identified as serotype 4 (HAdV-4) by neutralization test [
A straightforward typing system might help improve epidemiological investigations. Here
we have presented a rapid, generic, multigene typing system for human adenoviruses that can
characterize three main deterministic regions of HAdV strains and thus distinguish the
nonrecombinant HAdV types, and most recombinant HAdV types also. Using this system, we
analyzed 281 HAdV clinical isolates, and could detect not only divergent strains of established
types, but also a new recombinant HAdV strain with a previously unpublished combination of
HAdV genomes. This strain was approved by the Human Adenovirus Working Group to
represent a novel genotype: HAdV-86.
S1 Table. Constitution of the polymerase chain reactions used.
S2 Table. Thermal profiles of the polymerase chain reactions used.
S3 Table. Rough typing results of Swedish human adenovirus clinical isolates based on
partial amino acid sequences.
S4 Table. Sequence read coverage and basic attributes of the completely sequenced
We are especially grateful to Dr. Bengt Lo?fgren of Sk?ne University Hospital for providing
valuable virus strains. The assemblies and phylogenetic calculations were performed using
resources provided by SNIC through Uppsala Multidisciplinary Centre for Advanced
Computational Science (UPPMAX) under Project SNIC 2017/7-285. We are grateful for the kind and
helpful support provided by the Human Adenovirus Working Group in the evaluation of the
new type candidate strains.
Conceptualization: Gy?z? La?szlo? Kaja?n, Niklas Arnberg.
Data curation: Gy?z? La?szlo? Kaja?n, Da?niel Bartha.
Formal analysis: Gy?z? La?szlo? Kaja?n, Da?niel Bartha.
Funding acquisition: Niklas Arnberg.
Investigation: Gy?z? La?szlo? Kaja?n, Agnieszka Lipiec.
Methodology: Gy?z? La?szlo? Kaja?n.
Resources: Annika Allard.
11 / 15
Supervision: Niklas Arnberg.
Visualization: Gy?z? La?szlo? Kaja?n.
Writing ? original draft: Gy?z? La?szlo? Kaja?n.
Writing ? review & editing: Gy?z? La?szlo? Kaja?n, Annika Allard, Niklas Arnberg.
12 / 15
72. https://doi.org/10.1128/JVI.06165-11 PMID: 22072746; PubMed Central PMCID:
13 / 15
of intersubtype recombination. J Virol. 1999; 73(1):152?60. PMID: 9847317; PubMed Central PMCID:
14 / 15
(1):10. Epub 2018/02/07. https://doi.org/10.1038/s41426-017-0004-y PMID: 29410402; PubMed
Central PMCID: PMCPMC5837155.
1. Harrach B , Benk? M , Both GW , Brown M , Davison AJ , Echavarr??a M, et al. Family Adenoviridae . In: King AMQ , Lefkowitz E , Adams MJ , Carstens EB , editors. Virus Taxonomy: IXth Report of the International Committee on Taxonomy of Viruses . San Diego: Elsevier; 2011 . p. 125 - 41 .
2. Mei Y-F , Harrach B , Wadell G. Mastadenovirus . In: Tidona CA, Darai G , editors. The Springer Index of Viruses . 2 ed. New York: Springer-Verlag; 2011 . p. 33 - 48 .
3. Madisch I , Wo?lfel R , Harste G , Pommer H , Heim A . Molecular identification of adenovirus sequences: a rapid scheme for early typing of human adenoviruses in diagnostic samples of immunocompetent and immunodeficient patients . J Med Virol . 2006 ; 78 ( 9 ): 1210 - 7 . https://doi.org/10.1002/jmv.20683 PMID: 16847954 .
4. Madisch I , Harste G , Pommer H , Heim A . Phylogenetic analysis of the main neutralization and hemagglutination determinants of all human adenovirus prototypes as a basis for molecular classification and taxonomy . J Virol . 2005 ; 79 ( 24 ): 15265 - 76 . https://doi.org/10.1128/JVI.79.24. 15265 - 15276 . 2005 PMID: 16306598; PubMed Central PMCID : PMCPMC1316018 .
5. Allard A , Albinsson B , Wadell G . Rapid typing of human adenoviruses by a general PCR combined with restriction endonuclease analysis . J Clin Microbiol . 2001 ; 39 ( 2 ): 498 - 505 . https://doi.org/10.1128/JCM. 39.2. 498 - 505 . 2001 PMID: 11158096; PubMed Central PMCID : PMCPMC87765 .
6. Kaja?n GL , Kajon AE , Pinto AC , Bartha D , Arnberg N. The complete genome sequence of human adenovirus 84, a highly recombinant new Human mastadenovirus D type with a unique fiber gene . Virus Res . 2017 ; 242 : 79 - 84 . Epub 2017/09/18. https://doi.org/10.1016/j.virusres. 2017 . 09 .012 PMID: 28923509 .
7. Alonso-Padilla J , Papp T , Kaja?n GL , Benk? M , Havenga M , Lemckert A , et al. Development of Novel Adenoviral Vectors to Overcome Challenges Observed With HAdV-5-based Constructs . Mol Ther . 2016 ; 24 ( 1 ): 6 - 16 . https://doi.org/10.1038/mt. 2015 .194 PMID: 26478249; PubMed Central PMCID : PMCPMC4754553 .
8. Majhen D , Calderon H , Chandra N , Fajardo CA , Rajan A , Alemany R , et al. Adenovirus-based vaccines for fighting infectious diseases and cancer: progress in the field . Hum Gene Ther . 2014 ; 25 ( 4 ): 301 - 17 . https://doi.org/10.1089/hum. 2013 .235 PMID: 24580050 .
9. Baden LR , Karita E , Mutua G , Bekker LG , Gray G , Page-Shipp L , et al. Assessment of the Safety and Immunogenicity of 2 Novel Vaccine Platforms for HIV-1 Prevention: A Randomized Trial . Ann Intern Med . 2016 ; 164 ( 5 ): 313 - 22 . https://doi.org/10.7326/M15-0880 PMID: 26833336.
10. Zhang Q , Seto D. Chimpanzee Adenovirus Vector Ebola Vaccine-Preliminary Report . N Engl J Med . 2015 ; 373 ( 8 ): 775 - 6 . https://doi.org/10.1056/NEJMc1505499 PMID: 26287858 .
11. Duffy MR , Alonso-Padilla J , John L , Chandra N , Khan S , Ballmann MZ , et al. Generation and characterization of a novel candidate gene therapy and vaccination vector based on human species D adenovirus type 56 . J Gen Virol . 2018 ; 99 ( 1 ): 135 - 47 . Epub 2017/11/20. https://doi.org/10.1099/jgv.0.000978 PMID: 29154744 .
12. Bradshaw AC , Baker AH . Gene therapy for cardiovascular disease: perspectives and potential . Vascul Pharmacol . 2013 ; 58 ( 3 ): 174 - 81 . Epub 2012/11/07. https://doi.org/10.1016/j.vph. 2012 . 10 .008 PMID: 23142171 .
13. Nemerow GR , Stewart PL , Reddy VS . Structure of human adenovirus . Curr Opin Virol . 2012 ; 2 ( 2 ): 115 - 21 . https://doi.org/10.1016/j.coviro. 2011 . 12 .008 PMID: 22482707; PubMed Central PMCID : PMCPMC3322384 .
14. Norrby E. The structural and functional diversity of Adenovirus capsid components . J Gen Virol . 1969 ; 5 ( 2 ): 221 - 36 . https://doi.org/10.1099/ 0022 -1317-5-2-221 PMID: 4899805 .
15. Bradley RR , Lynch DM , Iampietro MJ , Borducchi EN , Barouch DH . Adenovirus serotype 5 neutralizing antibodies target both hexon and fiber following vaccination and natural infection . J Virol . 2012 ; 86 ( 1 ): 625 - 9 . https://doi.org/10.1128/JVI.06254-11 PMID: 22072776; PubMed Central PMCID : PMCPMC3255879 .
16. Bradley RR , Maxfield LF , Lynch DM , Iampietro MJ , Borducchi EN , Barouch DH . Adenovirus serotype 5- specific neutralizing antibodies target multiple hexon hypervariable regions . J Virol . 2012 ; 86 ( 2 ): 1267 -
17. Ma J , Duffy MR , Deng L , Dakin RS , Uil T , Custers J , et al. Manipulating adenovirus hexon hypervariable loops dictates immune neutralisation and coagulation factor X-dependent cell interaction in vitro and in vivo . PLoS Pathog . 2015 ; 11 ( 2 ):e1004673. https://doi.org/10.1371/journal.ppat.1004673 PMID: 25658827; PubMed Central PMCID : PMCPMC4450073 .
18. Gall JG , Crystal RG , Falck-Pedersen E . Construction and characterization of hexon-chimeric adenoviruses: specification of adenovirus serotype . J Virol . 1998 ; 72 ( 12 ): 10260 - 4 . PMID: 9811772; PubMed Central PMCID : PMCPMC110610 .
19. Seto D , Chodosh J , Brister JR , Jones MS . Using the whole-genome sequence to characterize and name human adenoviruses . J Virol . 2011 ; 85 ( 11 ): 5701 - 2 . https://doi.org/10.1128/JVI.00354-11 PMID: 21450823; PubMed Central PMCID : PMCPMC3094998 .
20. Xu W , McDonough MC , Erdman DD . Species-specific identification of human adenoviruses by a multiplex PCR assay . J Clin Microbiol . 2000 ; 38 ( 11 ): 4114 - 20 . PMID: 11060077; PubMed Central PMCID : PMCPMC87550 .
21. Lu X , Erdman DD . Molecular typing of human adenoviruses by PCR and sequencing of a partial region of the hexon gene . Arch Virol . 2006 ; 151 ( 8 ): 1587 - 602 . https://doi.org/10.1007/s00705-005-0722-7 PMID: 16502282 .
22. Denisov GA , Arehart AB , Curtin MD , inventors; Paracel, Inc. (Pasadena, CA), assignee. A system and method for improving the accuracy of DNA sequencing and error probability estimation through application of a mathematical model to the analysis of electropherograms . United States of America patent 6681186 . 2004 .
23. Huang X , Madan A . CAP3: A DNA sequence assembly program . Genome Res . 1999 ; 9 ( 9 ): 868 - 77 . PMID: 10508846; PubMed Central PMCID : PMCPMC310812 .
24. Muhire BM , Varsani A , Martin DP . SDT: a virus classification tool based on pairwise sequence alignment and identity calculation . PLoS One . 2014 ; 9 ( 9 ):e108277. https://doi.org/10.1371/journal.pone. 0108277 PMID: 25259891; PubMed Central PMCID : PMCPMC4178126 .
25. Horwitz MS . Adenoviruses. In: Fields BN , Knipe DM , editors. Fields virology . New York: Raven Press; 1990 . p. 1723 - 37 .
26. Zhou X , Robinson CM , Rajaiya J , Dehghan S , Seto D , Jones MS , et al. Analysis of human adenovirus type 19 associated with epidemic keratoconjunctivitis and its reclassification as adenovirus type 64 . Invest Ophthalmol Vis Sci . 2012 ; 53 ( 6 ): 2804 - 11 . Epub 2012/05/14. https://doi.org/10.1167/iovs.12- 9656 PMID: 22467570; PubMed Central PMCID : PMCPMC3367469 .
27. Kajon AE , Erdman DD . Assessment of genetic variability among subspecies B1 human adenoviruses for molecular epidemiology studies . Methods Mol Med . 2007 ; 131 : 335 - 55 . PMID: 17656793 .
28. Chevreux B , Wetter T , Suhai S , editors. Genome sequence assembly using trace signals and additional sequence information . German conference on bioinformatics; 1999: Heidelberg.
29. Kearse M , Moir R , Wilson A , Stones-Havas S , Cheung M , Sturrock S , et al. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data . Bioinformatics . 2012 ; 28 ( 12 ): 1647 - 9 . Epub 2012/04/27. https://doi.org/10.1093/bioinformatics/bts199 PMID: 22543367; PubMed Central PMCID : PMCPMC3371832 .
30. Yuan X , Qu Z , Wu X , Wang Y , Liu L , Wei F , et al. Molecular modeling and epitopes mapping of human adenovirus type 3 hexon protein . Vaccine . 2009 ; 27 ( 37 ): 5103 - 10 . Epub 2009/06/30. https://doi.org/10. 1016/j.vaccine. 2009 . 06 .041 PMID: 19573641 .
31. Katoh K , Standley DM . MAFFT multiple sequence alignment software version 7: improvements in performance and usability . Mol Biol Evol . 2013 ; 30 ( 4 ): 772 - 80 . Epub 2013/01/16. https://doi.org/10.1093/ molbev/mst010 PMID: 23329690; PubMed Central PMCID : PMCPMC3603318 .
32. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies . Bioinformatics . 2014 ; 30 ( 9 ): 1312 - 3 . https://doi.org/10.1093/bioinformatics/btu033 PMID: 24451623; PubMed Central PMCID : PMCPMC3998144 .
33. Talavera G , Castresana J . Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments . Syst Biol . 2007 ; 56 ( 4 ): 564 - 77 . https://doi.org/10. 1080/10635150701472164 PMID: 17654362 .
34. Kumar S , Stecher G , Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets . Mol Biol Evol . 2016 ; 33 ( 7 ): 1870 - 4 . Epub 2016/03/22. https://doi.org/10.1093/molbev/ msw054 PMID: 27004904 .
35. Lole KS , Bollinger RC , Paranjape RS , Gadkari D , Kulkarni SS , Novak NG , et al. Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence
36. Bodi K , Perera AG , Adams PS , Bintzler D , Dewar K , Grove DS , et al. Comparison of commercially available target enrichment methods for next-generation sequencing . J Biomol Tech . 2013 ; 24 ( 2 ): 73 - 86 . https://doi.org/10.7171/jbt.13-2402-002 PMID: 23814499; PubMed Central PMCID : PMCPMC3605921 .
37. Pauly M , Hoppe E , Mugisha L , Petrzelkova K , Akoua-Koffi C , Couacy-Hymann E , et al. High prevalence and diversity of species D adenoviruses (HAdV-D) in human populations of four Sub-Saharan countries . Virol J . 2014 ; 11 : 25 . https://doi.org/10.1186/ 1743 -422X- 11 -25 PMID: 24512686; PubMed Central PMCID : PMCPMC3928611 .
38. Adhikary AK , Inada T , Banik U , Numaga J , Okabe N. Identification of subgenus C adenoviruses by fiber-based multiplex PCR . J Clin Microbiol . 2004 ; 42 ( 2 ): 670 - 3 . https://doi.org/10.1128/JCM.42.2. 670 - 673 . 2004 PMID: 14766835; PubMed Central PMCID : PMCPMC344504 .
39. Ebner K , Rauch M , Preuner S , Lion T. Typing of human adenoviruses in specimens from immunosuppressed patients by PCR-fragment length analysis and real-time quantitative PCR . J Clin Microbiol . 2006 ; 44 ( 8 ): 2808 - 15 . https://doi.org/10.1128/JCM.00048-06 PMID: 16891496; PubMed Central PMCID : PMCPMC1594637 .
40. Kosulin K , Berkowitsch B , Lion T. Modified pan-adenovirus real-time PCR assay based on genome analysis of seventy HAdV types . J Clin Virol . 2016 ; 80 : 60 - 1 . Epub 2016/05/04. https://doi.org/10.1016/ j.jcv. 2016 . 05 .001 PMID: 27164448 .
41. Metzgar D , Osuna M , Kajon AE , Hawksworth AW , Irvine M , Russell KL . Abrupt emergence of diverse species B adenoviruses at US military recruit training centers . J Infect Dis . 2007 ; 196 ( 10 ): 1465 - 73 . https://doi.org/10.1086/522970 PMID: 18008225 .
42. Han G , Niu H , Zhao S , Zhu B , Wang C , Liu Y , et al. Identification and typing of respiratory adenoviruses in Guangzhou, Southern China using a rapid and simple method . Virol Sin . 2013 ; 28 ( 2 ): 103 - 8 . https:// doi.org/10.1007/s12250-013-3308-7 PMID: 23575732 .
43. Takeuchi S , Itoh N , Uchio E , Aoki K , Ohno S . Serotyping of adenoviruses on conjunctival scrapings by PCR and sequence analysis . J Clin Microbiol . 1999 ; 37 ( 6 ): 1839 - 45 . PMID: 10325334; PubMed Central PMCID : PMCPMC84965 .
44. Metzgar D , Osuna M , Yingst S , Rakha M , Earhart K , Elyan D , et al. PCR analysis of egyptian respiratory adenovirus isolates, including identification of species, serotypes, and coinfections . J Clin Microbiol . 2005 ; 43 ( 11 ): 5743 - 52 . https://doi.org/10.1128/JCM.43.11. 5743 - 5752 . 2005 PMID: 16272512; PubMed Central PMCID : PMCPMC1287779 .
45. McCarthy T , Lebeck MG , Capuano AW , Schnurr DP , Gray GC . Molecular typing of clinical adenovirus specimens by an algorithm which permits detection of adenovirus coinfections and intermediate adenovirus strains . J Clin Virol . 2009 ; 46 ( 1 ): 80 - 4 . https://doi.org/10.1016/j.jcv. 2009 . 06 .008 PMID: 19577957; PubMed Central PMCID : PMCPMC2846087 .
46. Sarantis H , Johnson G , Brown M , Petric M , Tellier R . Comprehensive detection and serotyping of human adenoviruses by PCR and sequencing . J Clin Microbiol . 2004 ; 42 ( 9 ): 3963 - 9 . https://doi.org/10. 1128/JCM.42.9. 3963 - 3969 . 2004 PMID: 15364976; PubMed Central PMCID : PMCPMC516336 .
47. Hage E , Gerd Liebert U , Bergs S , Ganzenmueller T , Heim A . Human mastadenovirus type 70: a novel, multiple recombinant species D mastadenovirus isolated from diarrhoeal faeces of a haematopoietic stem cell transplantation recipient . J Gen Virol . 2015 ; 96 ( 9 ): 2734 - 42 . https://doi.org/10.1099/vir.0. 000196 PMID: 26002300 .
48. Kajon AE , Lamson D , Shudt M , Oikonomopoulou Z , Fisher B , Klieger S , et al. Identification of a novel intertypic recombinant species D human adenovirus in a pediatric stem cell transplant recipient . J Clin Virol . 2014 ; 61 ( 4 ): 496 - 502 . https://doi.org/10.1016/j.jcv. 2014 . 09 .009 PMID: 25449172; PubMed Central PMCID : PMCPMC4276453 .
49. Esposito S , Zampiero A , Bianchini S , Mori A , Scala A , Tagliabue C , et al. Epidemiology and Clinical Characteristics of Respiratory Infections Due to Adenovirus in Children Living in Milan, Italy, during 2013 and 2014 . PLoS One . 2016 ; 11 ( 4 ):e0152375. https://doi.org/10.1371/journal.pone.0152375 PMID: 27045588; PubMed Central PMCID : PMCPMC4821614 .
50. Pring-Akerblom P , Adrian T . Type - and group-specific polymerase chain reaction for adenovirus detection . Res Virol . 1994 ; 145 ( 1 ): 25 - 35 . PMID: 8023012 .
51. Seto D , Jones MS , Dyer DW , Chodosh J . Characterizing, typing, and naming human adenovirus type 55 in the era of whole genome data . J Clin Virol . 2013 ; 58 ( 4 ): 741 - 2 . https://doi.org/10.1016/j.jcv. 2013 . 09 .025 PMID: 24161575 .
52. Ismail AM , Cui T , Dommaraju K , Singh G , Dehghan S , Seto J , et al. Genomic analysis of a large set of currently-and historically-important human adenovirus pathogens . Emerg Microbes Infect . 2018 ; 7
53. Robinson CM , Singh G , Lee JY , Dehghan S , Rajaiya J , Liu EB , et al. Molecular evolution of human adenoviruses . Sci Rep . 2013 ; 3:1812 . https://doi.org/10.1038/srep01812 PMID: 23657240; PubMed Central PMCID : PMCPMC3648800 .
54. Gonzalez G , Koyanagi KO , Aoki K , Watanabe H. Interregional Coevolution Analysis Revealing Functional and Structural Interrelatedness between Different Genomic Regions in Human Mastadenovirus D . J Virol. 2015 ; 89 ( 12 ): 6209 - 17 . https://doi.org/10.1128/JVI.00515-15 PMID: 25833048; PubMed Central PMCID : PMCPMC4474320 .
Walsh MP , Chintakuntlawar A , Robinson CM , Madisch I , Harrach B , Hudson NR , et al. Evidence of molecular evolution driven by recombination events influencing tropism in a novel human adenovirus that causes epidemic keratoconjunctivitis . PLoS One . 2009 ; 4 ( 6 ):e5635. https://doi.org/10.1371/journal. pone.0005635 PMID: 19492050; PubMed Central PMCID : PMCPMC2685984 .
56. Sriwanna P , Chieochansin T , Vuthitanachot C , Vuthitanachot V , Theamboonlers A , Poovorawan Y. Molecular characterization of human adenovirus infection in Thailand, 2009 - 2012 . Virol J. 2013 ; 10 : 193 . https://doi.org/10.1186/ 1743 -422X- 10 -193 PMID: 23758792; PubMed Central PMCID : PMCPMC3693972 .
57. Al Qurashi YM , Guiver M , Cooper RJ . Sequence typing of adenovirus from samples from hematological stem cell transplant recipients . J Med Virol . 2011 ; 83 ( 11 ): 1951 - 8 . https://doi.org/10.1002/jmv.22204 PMID: 21915870 .
58. Selvaraju SB , Kovac M , Dickson LM , Kajon AE , Selvarangan R . Molecular epidemiology and clinical presentation of human adenovirus infections in Kansas City children . J Clin Virol . 2011 ; 51 ( 2 ): 126 - 31 . https://doi.org/10.1016/j.jcv. 2011 . 02 .014 PMID: 21440492 .
59. Barrero PR , Valinotto LE , Tittarelli E , Mistchenko AS . Molecular typing of adenoviruses in pediatric respiratory infections in Buenos Aires, Argentina ( 1999 - 2010 ). J Clin Virol . 2012 ; 53 ( 2 ): 145 - 50 . https:// doi.org/10.1016/j.jcv. 2011 . 11 .001 PMID: 22138300 .
60. Singh G , Robinson CM , Dehghan S , Jones MS , Dyer DW , Seto D , et al. Homologous recombination in E3 genes of human adenovirus species D . J Virol. 2013 ; 87 ( 22 ): 12481 - 8 . Epub 2013/09/11. https://doi. org/10.1128/JVI.01927-13 PMID: 24027303; PubMed Central PMCID : PMCPMC3807923 .
61. Dehghan S , Seto J , Liu EB , Walsh MP , Dyer DW , Chodosh J , et al. Computational analysis of four human adenovirus type 4 genomes reveals molecular evolution through two interspecies recombination events . Virology . 2013 ; 443 ( 2 ): 197 - 207 . Epub 2013/06/10. https://doi.org/10.1016/j.virol. 2013 . 05 .014 PMID: 23763770; PubMed Central PMCID : PMCPMC3779658 .
62. Li QG , Wadell G. The degree of genetic variability among adenovirus type 4 strains isolated from man and chimpanzee . Arch Virol . 1988 ; 101 ( 1-2 ): 65 - 77 . PMID: 2843149 .