Population structure of the dengue viruses, Aragua, Venezuela, 2006–2007. Insights into dengue evolution under hyperendemic transmission

Abstract

During the past three decades there has been a notable increase in dengue disease severity in Venezuela. Nevertheless, the population structure of the viruses being transmitted in this country is not well understood. Here, we present a molecular epidemiological study on dengue viruses (DENV) circulating in Aragua State, Venezuela during 2006-2007. Twenty-one DENV full-length genomes representing all of the four serotypes were amplified and sequenced directly from the serum samples. Notably, only DENV-2 was associated with severe disease. Phylogenetic trees constructed using Bayesian methods indicated that only one genotype was circulating for each serotype. However, extensive viral genetic diversity was found in DENV isolated from the same area during the same period, indicating significant in situ evolution since the introduction of these genotypes. Collectively, the results suggest that the non-structural (NS) proteins may play an important role in DENV evolution, particularly NS1, NS2A and NS4B proteins. The phylogenetic data provide evidence to suggest that multiple introductions of DENV have occurred from the Latin American region into Venezuela and vice versa. The implications of the significant viral genetic diversity generated during hyperendemic transmission, particularly in NS protein are discussed and considered in the context of future development and use of human monoclonal antibodies as antivirals and tetravalent vaccines.

Full text

Population structure of the dengue viruses, Aragua, Venezuela, 2006–2007.
Insights into dengue evolution under hyperendemic transmission
Rosmari Rodriguez-Roche
a,
⇑
, Elci Villegas
b
, Shelley Cook
c
, Pauline A.W. Poh Kim
d
, Yoandri Hinojosa
a
,
Delfina Rosario
a
, Iris Villalobos
e
, Herminia Bendezu
b
, Martin L. Hibberd
d
, Maria G. Guzman
a
a
‘‘Pedro Kouri’’ Tropical Medicine Institute, P.O. Box 601, Marianao 13, Havana, Cuba
b
Instituto Experimental ‘‘José Witremundo Torrealba’’, ULA, Trujillo, Venezuela
c
Natural History Museum, Cromwell Road, London SW7 5BD, United Kingdom
d
Genome Institute of Singapore, Singapore 138672, Singapore
e
Hospital Central de Maracay, Universidad de Carabobo, Venezuela
article info
Article history:
Received 7 October 2011
Received in revised form 8 December 2011
Accepted 10 December 2011
Available online 17 December 2011
Keywords:
Dengue
Evolution
Phylogeny
Venezuela
abstract
During the past three decades there has been a notable increase in dengue disease severity in Venezuela.
Nevertheless, the population structure of the viruses being transmitted in this country is not well under-
stood. Here, we present a molecular epidemiological study on dengue viruses (DENV) circulating in Ara-
gua State, Venezuela during 2006–2007. Twenty-one DENV full-length genomes representing all of the
four serotypes were amplified and sequenced directly from the serum samples. Notably, only DENV-2
was associated with severe disease. Phylogenetic trees constructed using Bayesian methods indicated
that only one genotype was circulating for each serotype. However, extensive viral genetic diversity
was found in DENV isolated from the same area during the same period, indicating significant in situ evo-
lution since the introduction of these genotypes. Collectively, the results suggest that the non-structural
(NS) proteins may play an important role in DENV evolution, particularly NS1, NS2A and NS4B proteins.
The phylogenetic data provide evidence to suggest that multiple introductions of DENV have occurred
from the Latin American region into Venezuela and vice versa. The implications of the significant viral
genetic diversity generated during hyperendemic transmission, particularly in NS protein are discussed
and considered in the context of future development and use of human monoclonal antibodies as
antivirals and tetravalent vaccines.
Ó2011 Elsevier B.V.
1. Introduction
DENV is a single-stranded, positive-sense RNA virus belonging
to the genus Flavivirus, family Flaviviridae. There are four antigen-
ically distinct serotypes (DENV-1 to -4), all of which can cause ill-
ness. Dengue has a wide spectrum of clinical presentations, often
with unpredictable clinical evolution and outcome. While most pa-
tients recover following a self-limiting non-severe clinical course, a
small proportion progress to severe disease, mostly characterised
by plasma leakage with or without haemorrhage (WHO/TDR,
2009). In many tropical and subtropical countries DENV virus con-
stitutes a major public health problem. During the past 30 years or
so, in the Americas, dengue disease has increased dramatically.
Over 4.5 million cases were reported during 2000–2007, compared
with approximately 1 million cases previously reported in the
1980s. Likewise, the number of DHF cases increased over time from
13,398 (0.2/100,000) during the 1980s, to 111,724 (1.7/100,000)
during 2000–2007. From 1980 to 2007, Brazil reported the major-
ity of dengue cases (54.5%). However, Venezuela reported the high-
est number of DHF cases (35.1%) during the same period (San
Martin et al., 2010).
The first dengue epidemic was reported in Venezuela in 1964,
which lasted until 1967 (PAHO, 1979). During this time, 23 deaths
were attributed to DENV, but no investigations were carried out to
confirm the occurrence of DHF/DSS as the cause of death. After-
wards, an outbreak due to DENV-2 was reported in 1969. Subse-
quently from 1971 until 1977, the small number of reported
indicated that DENV activity was low (PAHO, 1979). However,
DENV-1 was introduced in 1977, firstly into Jamaica and Cuba
and 1 year later, in Venezuela and Puerto Rico causing massive epi-
demics of dengue (PAHO, 1979). During the following 4 years,
DENV-1 spread throughout the Caribbean Islands, Mexico, Texas,
Central America and South America. In 1981, DENV-4 was intro-
duced into the Americas but did not cause a major epidemic in
1567-1348 Ó2011 Elsevier B.V.
doi:10.1016/j.meegid.2011.12.005
⇑
Corresponding author. Address: Department of Virology, PAHO/WHO Collabo-
rating Center for the Study of Dengue and its Vector, ‘‘Pedro Kourí’’ Tropical
Medicine Institute, P.O. Box 601, Marianao 13, Havana, Cuba. Tel.: +53 7 2020450;
fax: +53 7 2046051.
E-mail address: rosmari@ipk.sld.cu (R. Rodriguez-Roche).
Infection, Genetics and Evolution 12 (2012) 332–344
Contents lists available at SciVerse ScienceDirect
Infection, Genetics and Evolution
journal homepage: www.elsevier.com/locate/meegid
Open access under CC BY license.
Open access under CC BY license.

Venezuela. Also in 1981, a DENV-2 strain of Asian origin was intro-
duced into Cuba, causing the first epidemic of DHF in the Americas
(Guzman et al., 1995; Kouri et al., 1986, 1987). However, a second
large epidemic of DHF occurred in Venezuela in 1989–1990 (PAHO,
1990). At the same time DENV-1, DENV-2 and DENV-4 were circu-
lating although DENV-2 virus was associated with most fatal cases
(Gubler and Meltzer, 1999). DENV-3 re-appeared in the Latin
American region in 1994 after an absence of 17 years (Guzman,
1995). The virus was detected almost simultaneously in Panama,
causing a small outbreak of classic DF, and in Nicaragua, where it
was associated with a nationwide epidemic of DF/DHF. After the
spread of the virus to the Latin American region this serotype
re-appeared in Venezuela in 2000 causing the largest dengue epi-
demic in Venezuela since 1989 (Uzcategui et al., 2003). Currently,
like in most Latin American countries, dengue is endemic in
Venezuela but the population structure of the viruses being trans-
mitted is not well understood. Phylogenetics has enhanced our
understanding of DENV population dynamics and sizes at various
stages of infection and transmission and this information actually
improves our ability to predict DENV emergence (Weaver and
Vasilakis, 2009). Therefore, the present investigation was aimed
at characterisation of the DENV serotypes and genotypes circulat-
ing in Aragua State, Venezuela for the purposes of understanding
viral epidemiology and to provide useful information for the
potential development and use of antivirals and vaccines.
2. Material and methods
2.1. Patient enrolment and blood sample collection
Blood samples from suspected dengue cases were collected at
the Maracay Central Hospital, Venezuela, during the period
November 2006–April 2007 as part of the DENCO project: Towards
successful dengue prevention and control (Alexander et al., 2011).
Written informed consent was obtained from all patients or guard-
ians. The study received approval from the Ethics Committee of
Experimental Institute ‘‘José Witremundo Torrealba’’, Ethics Com-
mittee of ‘‘Pedro Kouri’’ Tropical Medicine Institute and from the
World Health Organization Research Ethics Review Committee
(number A70175).
2.2. Diagnostic tests
After separation from red blood cells, acute-phase serum sam-
ples were tested for DENV RNA using a nested reverse transcrip-
tase–polymerase chain reaction (RT-PCR) assay that targets the
capsid-premembrane region and allows both DENV detection and
serotyping (Lanciotti et al., 1992). Dengue positive samples were
studied using real-time RT-PCR to measure viral titre using a sero-
type-specific assay that has been described previously (Laue et al.,
1999). Viral isolation was also attempted by inoculation onto Aedes
albopictus C6/36 cells (Rodriguez-Roche et al., 2000). Serotyping
after viral isolation was performed by indirect immunofluores-
cence using monoclonal antibodies (Henchal et al., 1983).
2.3. Full-length viral genome amplification
Real-time PCR and/or viral isolation positive samples were pro-
cessed in order to amplify. The earliest acute sample of each pa-
tient enrolled in the study was utilised for sequencing purpose.
Briefly, viral RNA was extracted from 140
l
L of serum sample
using the QIAamp viral RNA mini kit (Qiagen, Germany). cDNA
was synthesized using the SuperScript III First-Strand Synthesis
System (Invitrogen, USA) using specific serotype primers comple-
mentary to the 3
0
UTR as described previously (Christenbury
et al., 2010). An aliquot of 3
l
l cDNA was subjected to PCR using
the Expand High Fidelity PCR System (Roche Applied Science, Ger-
many) according to the manufacturer’s instructions. Five pairs of
primers for each serotype were utilised, designed to obtain five
overlapped fragments (F1–F5) covering the complete genome of
the viruses (Christenbury et al., 2010).
2.4. Nucleotide sequencing
PCR products were purified using the QIAquick PCR Purification
Kit (Qiagen, USA). Direct sequencing of these products was carried
out using an Applied Biosystems BigDye ddNTP capillary sequencer
as described previously (Schreiber et al., 2009). The chromato-
grams from capillary sequencing were assembled into a specimen
consensus sequence using SeqScape version 2.5 (Applied Biosys-
tems). The minimum fold-coverage of the sequences was at least
2, but in average it was 3. The nucleotide sequences reported
in this study are available in GenBank ID: HQ332170–HQ332190.
2.5. Sequence analysis
Full polyprotein nucleotide sequences of each dengue serotype
obtained in the present study were aligned using ClustalX (Thomp-
son et al., 1997) together with relevant sequences retrieved from
GenBank (available from the authors on request) such that repre-
sentative sequences from all the known DENV genotypes were
present. From the initial data set, identical sequences and known
recombinant sequences (as published by the authors) were re-
moved from the alignments. This produced a total data set of
101 sequences for DENV-1 10176 nucleotides in length, 89 se-
quences for DENV-2 10173 nucleotides in length, 95 sequences
for DENV-3 10170 nucleotides in length and 77 sequences for
DENV-4 10161 nucleotides in length. Maximum likelihood (ML)
phylogenetic trees were estimated using the general time-revers-
ible model (GTR) of nucleotide substitution, with the GTR substitu-
tion matrix, the base composition, the gamma distribution of
among-site rate variation, and the proportion of invariant sites
all estimated from the data using Modeltest (Posada and Crandall,
1998). To assess the robustness of particular phylogenetic group-
ings, a bootstrap analysis was undertaken using 1000 replicate
neighbour-joining trees using the ML substitution matrix de-
scribed above. All analyses were performed using PAUP (Swofford,
2003).
Phylogenetic analyses were also performed using Bayesian
analysis in MrBayes v3.1.2 (Huelsenbeck and Ronquist, 2001), with
a minimum of 20 million generations and a burn-in of 10%. Station-
ary was assessed at effective sample size (ESS > 400) using Tracer
v1.4.1 (part of the BEAST package) (Drummond and Rambaut,
2007).
All Bioinformatic analyses were carried out on the freely avail-
able Bioportal: www.bioportal.uio.no.
3. Results
From 50 acute-phase sera, 31 samples yielded viral isolates and/
or were positive by RT-PCR tests (10 DENV-1, 10 DENV-2, 2 DENV-3
and 9 DENV-4), confirming the co-circulation of the four DENV
serotypes in Aragua State, Venezuela. All positive samples were
subjected to Reverse Transcription and PCR for full-length genome
amplification using the long PCR system mentioned above. Twenty-
one of 31 DENV were fully amplified directly from sera (7 DENV-1, 7
DENV-2, 2 DENV-3 and 5 DENV-4) (Table 1). The viral titre deter-
mined for each serum sample had a significant influence on the
likelihood of attaining a genome length sequence. Real time RT-
PCR positive samples ranking from 12,880 to 8 PFU/mL produced
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 333

positive long PCR products useful for sequencing. However, 10 sam-
ples with less than 4 PFU/mL resulted in unsuccessful long PCR
amplification. The low viral titre observed in these patients could
be due to the time of sample collection >3 days following fever on-
set, presumably in these cases viral clearance through an adequate
immune response to the infection was responsible for the mild dis-
ease observed amongst this group.
The analysis of the full-length viral sequences revealed the exis-
tence of a single genotype for each serotype in Aragua State during
the period of study. However, the extent of genetic variability ob-
served amongst viruses of the same genotype collected over a few
months of study was surprising. The most variable serotype was
DENV-1; although two isolates of this subgroup (VE_61059_2006
and VE_61060_2006), which were obtained from two members
of the same family (with fever onset only one day apart), had iden-
tical sequences. Comparison with the five remaining DENV-1 re-
vealed 2% of variable sites. On the other hand, comparison
among the seven DENV-2 sequences and the five DENV-4 se-
quences revealed 1.3% of variable sites in both serotypes. No differ-
ences were found between the only two DENV-3 studied herein,
which is most likely due to the limited number of DENV-3 isolates
recovered.
Comparison of the unique DENV-3 sequence obtained in the
present study with 12 full-length DENV-3 Venezuelan sequences
available in Genbank, corresponding to isolates from the same
locality (Aragua) and period (2006–2007) of study revealed up to
2.1% of divergence at the nucleotide and 1% at the amino acid level,
respectively. A similar analysis was carried out for the other sero-
types using full-length Venezuelan sequences available in Gen-
bank, corresponding to isolates from the same locality and period
i.e. 18 DENV-1, 8 DENV-2 and 22 DENV-4 sequences. Data for
DENV-1 revealed up to 2.3% of divergence in nucleotide sequences
and 0.9% at the amino acid level, 2.2% and 1% for DENV-2, 1% and
0.3% for DENV-4, respectively. Notably, DENV-4 showed lower var-
iability despite the fact that a larger number of sequences were
available for this serotype.
In general, for the four serotypes, most non-synonymous muta-
tions produced conservative amino acid changes. Notably, these
mutations were generally unique for particular isolates, suggesting
that they were not fixed within the population during the time of
study. However, significantly larger sample sizes will be required
to verify this observation. Nonetheless, the comparison of Venezue-
lan DENV sequences obtained in this study and those corresponding
to isolates from the same locality (Aragua) and period (2006–2007)
available in Genbank with globally distributed DENV sequences,
revealed specific amino acid changes in the Venezuelan isolates that
appeared to be fixed after the introduction of particular genotypes
into this country as well as specific amino acidic changes that
differentiate Latin-American isolates from other geographically
distant viruses but corresponding to the same genotype. Most of
these changes occurred in the non-structural genes (Table 2).
ML and Bayesian phylogenetic trees were constructed for each
dengue serotype. Since robust trees with very similar topologies
were obtained by both methods, only Bayesian phylogenies are
shown herein (Figs. 1–4). It should be noted that posterior proba-
bilities obtained via Bayesian methods, and neighbour-joining
bootstrap replicate values (not shown), were similarly high for all
major nodes.
Table 1
Data corresponding to 21 Venezuelan samples sequenced in the study.
Sample number Age (years) Date of fever onset Days after fever onset Serotype Clinical classification
a
Type of infection
b
Viral titre (PFU
c
/mL)
61006-1 19 05/11/2006 3 DENV-1 Non-severe S 57.5
61059-1 7 29/11/2006 3 DENV-1 Non-severe S 39.1
61060-1 4 30/11/2006 2 DENV-1 Non-severe S 8994
61063-1 3 02/12/2006 2 DENV-1 Non-severe P 25.9
61068-1 7 09/12/2006 3 DENV-1 Non-severe S 357.7
61081-1 11 20/01/2007 3 DENV-1 Non-severe P 18.3
61084-1 9 22/01/2007 3 DENV-1 Non-severe S 273.7
61069-1 38 10/12/2006 3 DENV-2 Non-severe S 35.6
61082-1 27 20/01/2007 4 DENV-2 Non-severe S 1578
61095-1 <1 08/02/2007 4 DENV-2 Severe P 6082
61115-1 21 17/03/2007 2 DENV-2 Severe S 942.7
61133-1 52 04/04/2007 3 DENV-2 Severe S 2277
61136-1 5 08/04/2007 1 DENV-2 Non-severe S 12890
61154-1 15 27/04/2007 3 DENV-2 Non-severe S 33.47
61035-1 51 20/11/2006 2 DENV-3 Non-severe S 280
61051-1 15 25/11/2006 3 DENV-3 Non-severe S 212.3
61013-1 12 06/11/2006 4 DENV-4 Non-severe S 8.04
61027-1 51 14/11/2006 2 DENV-4 Non-severe S 4819
61054-1 12 25/11/2006 4 DENV-4 Non-severe S 49.3
61073-1 29 02/01/2007 3 DENV-4 Non-severe S 1035
61110-1 14 01/03/2007 2 DENV-4 Non-severe S 5284
a
Clinical classification is shown by the New WHO/TDR Guidelines published in 2009.
b
S: secondary infection, P: primary infection.
c
PFU: plaque formation units.
Table 2
Summary of particular non-conserved amino acid replacements observed in recent
Venezuelan isolates.
Serotype Proteins
E NS1 NS2A NS4B NS5
DENV-1 D146G
a
H18Y
a
DENV-2 I93T
b
A109T
b
S24P
c
T264I
c
Q164L/R
a,b
DENV-3 K4E
a
S229A
d
P365S
d
K371R
d
E374G
d
R389K
d
R422K
d
E429D
d
DENV-4 A222T
a
D290N
a
S354A
a
a
Amino acid replacements in Venezuelan isolates unusual for global sequences.
b
Amino acid replacements in recent Venezuelan isolates compared with previ-
ous circulating lineages.
c
Amino acid replacements in Venezuelan isolates from the same period (2006–
2007).
d
Amino acid replacements in Latin American isolates compared with others
isolates corresponding to genotype III.
334 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344

Fig. 1. Bayesian phylogeny of DENV-1 polyprotein nucleotide data set, including Venezuelan isolates from 2006 to 2007. For clarity, posterior probabilities obtained via
Bayesian methods are shown for the main clades only. Bootstrap support values obtained via 1000 neighbour-joining replicates were similarly high for these nodes. All
horizontal branch lengths are drawn to scale; bar, 0.02 substitutions per site. The tree is midpoint-rooted for purposes of clarity only. The asterisks (⁄) indicate those DENV
sequences that are published for the first time in the current study.
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 335

Fig. 2. Bayesian phylogeny of the DENV-2 polyprotein nucleotide data set, including Venezuelan isolates from 2006 to 2007. For clarity, posterior probabilities obtained via
Bayesian methods are shown for the main clades only. Bootstrap support values obtained via 1000 neighbour-joining replicates were similarly high for these nodes. All
horizontal branch lengths are drawn to scale; bar, 0.02 substitutions per site. The tree is midpoint-rooted for purposes of clarity only. The asterisks (⁄) indicate those DENV
sequences that are published for the first time in the current study.
336 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344

Fig. 3. Bayesian phylogeny of the DENV-3 polyprotein nucleotide data set, including Venezuelan isolates from 2006 to 2007. For clarity, posterior probabilities obtained via
Bayesian methods are shown for the main clades only. Bootstrap support values obtained via 1000 neighbour-joining replicates were similarly high for these nodes. All
horizontal branch lengths are drawn to scale; bar, 0.3 substitutions per site. The tree is midpoint-rooted for purposes of clarity only. The asterisks(⁄) indicate those DENV
sequences that are published for the first time in the current study.
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 337

Fig. 4. Bayesian phylogeny of the DENV-4 polyprotein nucleotide data set, including Venezuelan isolates from 2006 to 2007. For clarity, posterior probabilities obtained via
Bayesian methods are shown for the main clades only. Bootstrap support values obtained via 1000 neighbour-joining replicates were similarly high for these nodes. All
horizontal branch lengths are drawn to scale; bar, 0.4 substitutions per site. The tree is midpoint-rooted for purposes of clarity only. The asterisks(⁄) indicate those DENV
sequences that are published for the first time in the current study.
338 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344

3.1. DENV-1
The Bayesian phylogeny obtained for the DENV-1 data set (Fig. 1)
confirms previous results concerning the genetic relatedness of the
Venezuelan isolates with other Latin American isolates grouping
within a single genotype (Goncalvez et al., 2002). The tree also re-
veals four sub-clusters within genotype III that include Venezuelan
isolates (A–D). However, it is striking that temporal distribution is
not observed; isolates from the 1990s are regularly mixed with the
most recent isolates. Sub-cluster A is mainly composed of sequences
corresponding to recent Venezuelan isolates (2005–2008) including
the strain VE_61063_2006 sequenced in this study, one strain from
Puerto Rico isolated in 2006 and two Venezuelan isolates from 1997.
Sub-cluster B mainly contained Venezuelan isolates (2004–2007)
including four isolates sequenced in this study (VE_61059_2006,
VE_61060_2006, VE_61068_2006 and VE_61084_2007), which
were closely related to Colombian isolates (2005–2007). Sub-cluster
C contained two recent Venezuelan isolates (2005 and 2008), a
closely related Venezuelan isolate from 1999 and two Colombian
isolates from 1998. Finally, sub-cluster D predominantly contained
Nicaraguan isolates (2004–2008) but included two Venezuelan
isolates from the present study and two Venezuelan isolates from
1997 to 1998. The isolates VE_61006_2006 (from the beginning of
November) and VE_61081_2007 (from the end of January, 2007),
closely related to strains circulating in Nicaragua (2004–2008),
showed 16 amino acid changes in common with the remaining
Venezuelan isolates sequenced in the present study i.e. E (1), NS1
(1), NS2A (2), NS2B (5), NS4A (1), NS4B (4) and NS5 (2).
In addition, it is noteworthy that with the exception of the two
closely related Nicaraguan isolates, all the Venezuelan isolates cor-
responding to the period 2006–2007 showed two non-conserva-
tive amino acid replacements which appears to be unusual for
DENV-1 of any genotype; i.e. D146G (polar acid to polar neutral)
in the NS1 protein and H18Y (polar basic to polar neutral) in the
NS4B protein. Particularly, the change at NS1 could be of relevance
since the motif D prefers generally to be on the surface of proteins
and it is quite frequently involved in protein active or binding sites.
On the contrary, the motif G can reside in parts of protein struc-
tures that are forbidden to all other amino acids. The uniqueness
of G also means that it can play a distinct functional role, such as
using its side chain-less backbone to bind to phosphates (Betts
and Russell, 2003).
3.2. DENV-2
The Bayesian tree constructed using the DENV-2 data set (Fig. 2)
shows previously identified genotypes (Asian Genotype 1, Asian
Genotype 2, Cosmopolitan Genotype, American and American/
Asian genotype) (Twiddy et al., 2002). Except for the oldest Vene-
zuelan strain (VE_BID_V3366_1987) included in this analysis,
which clustered with the American genotype, all the Venezuelan
isolates collected during the 1990s and more recently, clustered
within the American–Asian genotype suggesting that the former
genotype has been displaced. In general, geographical and tempo-
ral distribution in sub-clusters was observed within the American/
Asian genotype (e.g. Caribbean isolates were separated from main-
land isolates). Interestingly, two major Venezuelan phylogenetic
clusters can be distinguished within this genotype. Sub-cluster A
corresponds to isolates obtained during the 1990s and sub-cluster
B corresponds to Venezuelan isolates collected during the period
2006–2007 which are closely related to Colombian isolates from
2004 to 2005. The isolate CO_BID_3371_2005 is located at the base
of this group. Particularly, within sub-cluster B, the Venezuelan
isolates VE_61136_2007 and VE_61115_2007 appear slightly sepa-
rate from the rest of the Venezuelan isolates from 2006 to 2007.
These isolates have two non-conservative amino acid replace-
ments, T264I (polar neutral to non-polar neutral) in NS1 protein,
and S24P (polar neutral to non-polar neutral) in the region encod-
ing the NS4B protein.
Based on the polyprotein alignment, the more recent Venezue-
lan isolates differ from earlier isolates collected during the 1990s
by twenty amino acid replacements, nine of which appear to be
non-conservative and localised in the regions encoding the E (3),
NS1 (1), NS2A (2), NS4A (1) and NS5 (2) genes. Three of these nine
amino acid replacements, namely I93T (neutral non-polar to neu-
tral polar) in the NS1 protein, A109T (neutral non-polar to neutral
polar) and L164R (neutral non-polar to basic polar) both in the
NS2A protein, are absent in recent isolates grouped in the Ameri-
can/Asian Genotype with the exception of two strains from Colom-
bia, isolated in 2004 and 2005 which appear closely related to the
Venezuelan isolates from 2006 to 2007, suggesting that this line-
age may have been circulating earlier. The motif Q164 (neutral po-
lar) in the NS2A protein was conserved amongst DENV-2 of all
genotypes except for the Venezuelan strains circulating during
the 1990s which contained the motif L (neutral non-polar) at this
position and the Venezuelan 2006–2007 and Colombian 2004–
2005 isolates which contained R, a basic polar amino acid, which
may affect the hydrophobicity of the protein.
3.3. DENV-3
The Bayesian phylogeny obtained for the DENV-3 data set
(Fig. 3) shows that all Venezuelan isolates group within Genotype
III with other Latin American isolates from Nicaragua, Puerto Rico,
Brazil, Colombia and Mexico in agreement with previous studies
using the envelope gene (Ramirez et al., 2010; Uzcategui et al.,
2003). There is a Nicaraguan isolate from 1994 at the base of the
group (NI_BID_V2420_1994) coinciding with the year of introduc-
tion of this genotype into the Americas (Guzman, 1995; Guzman
et al., 1997; Rodriguez-Roche et al., 2005b). Despite the fact that
only 2 DENV-3 isolates were recovered from the samples collected
in the present study, other Venezuelan sequences published in
GenBank that correspond to the period of study, are very close
related.
Notably, the Venezuelan cluster includes isolates from 2001 to
2008, mixed only with Colombian isolates from 2002 to 2007, sug-
gesting an extensive virus interchange between Colombia and
Venezuela during this period of time. Interestingly, there are two
Venezuelan isolates outside this cluster, VE_BID_V911_2001 from
Caracas, closely related to Puerto Rican isolates (from 1999 and
2006) and VE_BID_V2971_2007 from Aragua grouping with a
Peruvian strain from 2002 and an Ecuadorian strain from 2000,
which are located at the base of the cluster that includes the most
recent Nicaraguan strains (2008–2009). Although this result sug-
gests that multiple introductions could occur and these lineages
could be circulating as non-predominant strains, their transmis-
sion has not been demonstrated here as only single isolates were
recovered.
Analysis of the DENV-3 polyprotein alignment revealed a un-
ique non-conservative amino acid change in Venezuelan isolates,
namely K4E (basic polar to acidic polar) at the N-terminal region
of the NS2A protein. Remarkably, this K motif at position 4 is con-
served globally amongst DENV-3 but with the exception of strains
VE_BID_V911_2001 and VE_BID_V2971_2007 which conserved K,
all the Venezuelan isolates from 2001 to 2007 have E at this posi-
tion. The amino acid K frequently plays an important role in pro-
tein structure. Firstly, it can be considered to be amphipathic
because the part of the side chain nearest to the backbone is long,
carbon containing and hydrophobic, whereas the end of the side
chain is positively charged (Betts and Russell, 2003). Therefore
the change K4E could be functionally important in the structure
of the NS2A protein.
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 339

Additional analysis revealed that seven amino acid replace-
ments in the NS5 protein differentiate the Latin American strains
from the remaining DENV-3 strains grouped in Genotype III (iso-
lated in Asia) and all the other DENV-3 genotypes. Importantly,
one of these changes (S229A) is localised in the core domain of
the S-adenosyl methionine-dependent methyltransferase (MTase).
The peptide
223
GNIVSSVN
232
encompassing this change (ie GNI-
VASVN) is otherwise conserved amongst all four DENV serotypes.
However, the change S229A is not observed in the Nicaraguan
strain NI_BID_V2420_1994, suggesting that the S229A change oc-
curred in situ following its introduction to the Americas in 1994.
However, the remaining six changes localised in the RNA-depen-
dent RNA polymerase (RdRp) at positions 365, 371, 374, 389, 422
and 429 were unique to the entire Latin American group in geno-
type III, including the 1994 Nicaraguan strain.
3.4. DENV-4
The Bayesian phylogeny obtained for the DENV-4 data set
(Fig. 4) shows that all Venezuelan isolates group within Genotype
II with other Latin American isolates. This genotype, known as the
Indonesian Genotype (relating to the origin of the virus), was intro-
duced into the Caribbean region in 1981 (Lanciotti et al., 1997). In
Fig. 4, the Dominican strain, isolated in 1981, is located at the base
of the Latin American group that is mainly represented by Puerto
Rican, Colombian and Venezuelan isolates, forming geographic
and temporally structured clusters. The Venezuelan isolates corre-
sponding to the present study are closely related to these Venezu-
elan viruses isolated during the same year and geographical
location, as previously published in Genbank. The phylogenetic
tree further indicates that in situ evolution of DENV-4 has occurred
in Venezuela, with differentiation into a number of distinct but co-
circulating lineages, rather than repeated introduction of new
strains from other localities. The Venezuelan isolates from Aragua
State are much more closely related to the Colombian strains from
different periods, as was observed in the DENV-3 phylogeny, con-
firming significant virus interchanges between these neighbouring
countries. The tree also shows dispersal of these lineages within
Venezuela, since a sequence retrieved from Genbank correspond-
ing to an isolate from Merida 2007 groups with other 2007 isolates
from geographically distant Aragua.
The polyprotein alignment of DENV-4 shows that isolates from
Venezuela 2007 and Colombia (2004–2005) have two amino acid
replacements in the E protein (A222T and S354A), which are unu-
sual for global DENV-4 strains. According to the complete genome
data set these changes were first identified in Venezuela in 2000.
Consequently, to examine variation at positions E222 and E354
in more detail, available E gene sequences were retrieved from
Genbank. Interestingly, similar to the Venezuelan, Peruvian and
Ecuadorian strains isolated in 2000 differ from the strains collected
during 2006–2008 at positions E222 and E354. Thus, based on the
data in hand the amino acid replacements A222T and S354A ap-
pear to have arisen very recently at positions, which are usually
invariant. Whether or not these changes impact on viral fitness
needs to be determined. However, the replacement at E222 is
localised in domain II in a surface exposed loop in the E protein
and the replacement at position E354 is localised in the C-terminal
domain III (amino acids 303–395) which contains residues associ-
ated with changes in host range, tropism, and virulence in different
flaviviruses (Rey et al., 1995).
In addition, the 2007 Venezuelan isolates appear to have an
amino acid replacement in the C-terminal region of the NS1
protein (D290N). In contrast residue D290 is conserved in all other
DENV-4 sequences independent of genotype. Consequently, all
isolates grouping within genotype II seem to display D at this
position, including the earlier strain Dominica/1981, as well as
the Venezuelan isolates collected from 1998 to 2001. Therefore,
this analysis suggests that this mutation in the NS1 gene may have
been fixed very recently. Although, it is striking that the sylvatic
strain MY_P75_215_1975 also has an N residue at NS1 290.
4. Discussion
The mechanistic basis for DHF has been a subject of study for
decades (Guzman, 2005; Mathew and Rothman, 2008; Tan and
Alonso, 2009). Although it is recognised that host and viral factors
are the primary determinants for the development of DHF in indi-
vidual patients, epidemiological and ecological conditions deter-
mine the occurrence of DHF epidemics (Guzman and Kouri,
2008). Studies derived from appropriate epidemiological settings
have contributed enormously to understanding the underlying
basis for the emergence of DHF. Risk factors for the development
of severe dengue disease include prior infection with a heterotypic
serotype, the strain of the infecting virus, age and gender and the
genetic background of the patient (Alvarez et al., 2006; Garcia
et al., 2010; Guzman et al., 2000, 2002a; Kouri et al., 1998, 1987;
Rodriguez-Roche et al., 2011; Sangkawibha et al., 1984; Sierra
et al., 2006, 2007; Vaughn et al., 2000).
Interestingly, on the basis of studies in Venezuela, one particu-
lar serotype has predominated during an epidemic period, namely
that which replaced the dominant serotype corresponding to the
previous period (Salas et al., 1998). Since 1989, all Venezuelan den-
gue outbreaks have commenced in the central region of the coun-
try, particularly in Aragua State, and then spread throughout the
entire country (Uzcategui et al., 2003). During the outbreak that
occurred from October 1989 to April 1990 which comprised over
6000 DHF cases and 73 deaths (PAHO, 1990), DENV-3 was the only
missing serotype and DENV-2 appeared to be associated most fre-
quently with fatal cases (Gubler, 1997; PAHO, 1990). A similar sit-
uation occurred from 1997 to 2000 when DENV-2 was once more
associated with increasing disease severity (Uzcategui et al., 2001).
During three different epidemics periods (1989, 1997, and 2006)
DENV-2 has been associated with increasing disease severity in
Venezuela, with the American/Asian genotype being responsible
for severe cases. Notably, the phylogenetic tree (Fig. 2) shows that
two different DENV-2 lineages have been circulating in Venezuela
during the last 20 years both grouped in the American/Asian geno-
type. One appears to be related to epidemics that occurred at the
end of the 1980s and the beginning of the 1990s and subsequently
at the end of the 1990s. The other lineage that has been circulating
more recently appears to have replaced the previous lineage. Basal
to the sub-cluster of recent Venezuelan isolates there are Colombian
isolates from 2004 to 2005, suggesting that the second introduction
could be from Colombia. Likewise, the Venezuelan isolates show
genetic relatedness with earlier isolates from the Caribbean, where
increasing severity due to DENV-2 was reported (Guzman et al.,
2002b; Rodriguez-Roche et al., 2005a). In the present study, the
relatively limited number of samples means that the hypothesis that
there is a correlation between disease severity, DENV serotype,
sequence of infection and viral genetic differences could not be
tested statistically. However, it is notable that according to
epidemiological data available all severe cases were only associated
with DENV-2; moreover, during the sequencing study three of seven
patients infected with this serotype had severe disease. One of these
patients (VE_61095_2007) was less than one-year-old and had
severe bleeding during primary infection. The other two patients
(VE_61115_2007 and VE_61133_2007) both adults, suffered
secondary infection in the sequence DENV-3/DENV-2 (Alvarez
et al., unpublished results). Because of an absence of bleeding, case
VE_61115_2007 was not considered as DHF according to the
WHO, 1997 strict classification (WHO, 1997). However, this patient
340 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344

had severe organ impairment and coma; consequently the infection
was classified as severe dengue according to the new dengue
classification (WHO/TDR, 2009). Case VE_61133_2007, which also
had severe organ impairment and coma, was considered as DHF
grade II according to the WHO, 1997 strict classification and severe
dengue according to the new guidelines TDR/WHO.
No particular viral changes were found that might be directly
related with the severe clinical picture observed during the course
of DENV-2 infections. For instance, the nucleotide sequence of
isolate VE_61095_2007, which caused severe bleeding during
dengue primary infection, is very similar to VE_61082_2007
obtained from a patient with non-severe disease. Likewise, isolate
VE_61115_2007 (from a severe case) is closely related to
VE_61136_2007 (from a non-severe case). Even so, it is important
to note that patient VE_61136_2007 showed the highest viral titre
among all the studied patients. Consequently, the amino acid
changes in the NS2A and NS4B observed only in these two isolates
collected late during the epidemic, could account for higher viral
fitness. Nevertheless, other non-viral factors could determine the
clinical outcome. Recently, it has been shown that a mutation in
the NS1 protein of DENV-2 may be associated with intra-epidemic
increasing disease severity during secondary infections but pre-
sumably this fittest virus can produce asymptomatic and mild dis-
ease in the vast majority of primary infected cases (Rodriguez-
Roche et al., 2011).
In general, DENV-1 is not frequently associated with a severe
disease outcome in the Americas. However, epidemiological data
support the possibility that this serotype may play an important
role in sensitising the population for a secondary infection (Alvarez
et al., 2006; Guzman and Kouri, 2008; Guzman et al., 2002b).
Apparently, after the introduction of DENV-1 into Latin America,
different lineages evolved independently to generate the diverse
genetic variants isolated during the last decade. During the present
study it has been demonstrated that different lineages have circu-
lated in Venezuela, presumably introduced from neighbouring
countries, where in situ evolution of Genotype III has occurred as
shown in the DENV-1 tree (Fig. 1). Previous studies using the re-
gion encoding the E gene have shown similar results regarding
the genetic relatedness of the DENV-1 American isolates (Goncal-
vez et al., 2002). The extensive genetic variability observed for
DENV-1 Venezuelan isolates should be examined considering that
the generation of new genetic variants could be linked to increased
disease severity. Despite the fact that only one genotype of DENV-1
has been circulating in the Americas since being introduced in
1977, an increasing incidence of severe dengue due to DENV-1
has been reported in Bolivia, 2008 and Argentina, 2009 compared
with data corresponding to previous circulation of this serotype
during 1999–2000 (Seijo, 2009). Unfortunately, full-length se-
quences of DENV-1 isolated from these outbreaks are not available.
Similarly, after the introduction DENV-3 (genotype III) in Latin
America in situ evolution has been observed. Particularly in Vene-
zuela there appears to be a number of distinct but co-circulating
lineages, rather than the repeated introduction of new strains from
other localities. Only two Venezuelan isolates VE_BID_V911_2001
and VE_BID_V2971_2007 do not fall with the Venezuelan 2000–
2008 cluster, indicating that two other genetically distant were
introduced although they either have limited circulation or have
been under sampled. Recent studies, based on a larger number of
envelope gene sequences of DENV-3 genotype III isolated in Vene-
zuela from 2001 to 2008, support this hypothesis since different
lineages within genotype III were found, suggesting that several
introduction events have occurred into this country. However,
the more recent Venezuelan isolates were located in the same clus-
ter as observed in the present study (Ramirez et al., 2010). None-
theless, it cannot be ignored that several introduced strains from
other parts of Latin America may have remained undetected in
our study due to sampling bias or gaps in national surveillance. In-
deed, in the present study DENV-3 was the least represented
amongst the four serotypes. In addition, most sequences included
in the analysis (isolates sequenced in this study plus the Venezue-
lan ones retrieved from Genbank) correspond mainly to Aragua
state. DENV-3 may have been displaced in recent years in Aragua
state, being the serotype identified the least frequently during
the period of the study. Nevertheless, it is important to consider
that this serotype could be involved in asymptomatic transmission.
For instance, the sequence DENV-2/DENV-3 has been statistically
associated with asymptomatic and non-severe disease in other
studies (Alvarez et al., 2006). More recently an increase in the cir-
culation of DENV-3 in Venezuela has been reported (Schmidt et al.,
2011).
Meanwhile, DENV-4 also shows a pattern of evolution charac-
terised by a number of distinct but co-circulating lineages that
do not seem to be associated with repeated introductions. As men-
tioned above, the pattern observed for DENV-3 and DENV-4 differs
from DENV-2 in which replacement of the circulating lineage after
a new introduction was associated with increasing proportion of
severe dengue cases (Uzcategui et al., 2001).
Most studies concerning the role of the virus in dengue patho-
genesis have been focussed on finding viral genetic variants asso-
ciated with severe disease. However, few studies have examined
the evolution of viruses that circulate extensively during hyperen-
demic transmission without causing severe disease. The results
presented here, including the four DENV serotypes, suggest the
non-structural proteins could play an important role in DENV evo-
lution, principally NS1, NS2A and NS4B. According to this particu-
lar epidemiological setting, changes in these proteins would either
be favourable or adverse in terms of viral fitness, since mutations
could be associated with severe disease (as possibly occurred with
DENV-2) or could be associated with non-severe disease due to the
appearance of naturally attenuated strains.
In the present study, the following non-conservative amino acid
replacements may be particularly noteworthy: D146G in the NS1
protein and H18Y in the NS4B protein, which are unusual for
DENV-1 of any genotype. According to previous reports, changes
in NS1 and NS4 proteins could be involved in viral attenuation
(Hanley et al., 2003; Kelly et al., 2009). On the contrary, recent
studies have demonstrated that mutations in the NS4B protein
may increase the efficiency of DENV replication. In addition, it
has been suggested that mutations in this protein may also be in-
volved in species tropism of DENV and modulate the balance of
efficient replication in mosquito and mammalian cells (Tajima
et al., 2011). Moreover, it has been shown that a single amino acid
in the non-structural NS4B protein namely L52F confers virulence
on DENV-2 in AG129 mice through enhancement of viral RNA syn-
thesis (Grant et al., 2011).
In the present study, the more divergent DENV-2 isolates re-
lated to severe dengue cases also showed non-conservative amino
acid replacements localised in the NS1 (T264I) and NS4B (S24P)
proteins.
The Venezuelan DENV-4 strains isolated during 2007 also
showed a particular amino acid change in the NS1 protein at the
C-terminal region, which differentiates them from DENV-4 strains
of any genotype. Previous studies have demonstrated that the fla-
vivirus NS1 region plays an essential role in the early steps of rep-
lication (Lindenbach and Rice, 1997; Mackenzie et al., 1996;
Muylaert et al., 1997). Deletion of the C-terminal region of DENV
NS1 protein abolishes anti-NS1-mediated platelet dysfunction
and bleeding tendency (Chen et al., 2009). Importantly, the C-ter-
minal region of the NS1 protein contains six cysteines, known to
play a role in protein folding. The motif NS1 290 that changed in
the DENV-4 Venezuelan isolates is located close to the cysteine 8
(position 291) of the NS1 protein (Wallis et al., 2004).
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 341

The analysis of the DENV-3 polyprotein alignment revealed that
there is a non-conservative amino acid change K4E in the N-termi-
nal region of NS2A protein. This is unusual for DENV-3 of any geno-
type, but it is specific to the more recent Venezuelan isolates from
2001 to 2007. The genetic variant containing the mutation K4E was
detected for the first time in Venezuela in 2001. Although the
selective force leading to the selection of this variant has not been
identified, it appears to have been fixed because this variant has
been circulating in Venezuela for at least 6-years.
The flavivirus NS2A protein is a small (231 amino acids), hydro-
phobic, multifunctional, membrane-associated protein involved in
RNA replication (Chambers et al., 1989; Mackenzie et al., 1998), in
host-antiviral interferon response (Liu et al., 2004, 2005, 2006; Mu-
noz-Jordan et al., 2003) and assembly/secretion of virus particles
(Kummerer and Rice, 2002; Leung et al., 2008; Liu et al., 2003).
In addition, NS2A and also NS4B protein may participate in the
modulation of vector competence (McElroy et al., 2006).
A retrospective phylogenetic study of events on the Southern
Pacific islands three decades ago, where severe dengue was de-
scribed in patients infected with the DENV-2 American genotype,
recorded attenuation of this virus following a series of outbreaks
involving non-synonymous mutations in the NS2A gene (Steel
et al., 2010). In contrast, during the latest and most severe DENV-
4 epidemic in Puerto Rico in 1998, viruses were distinguished by
three amino acid replacements in the NS2A protein (I14V, V54T
and P101T), which were fixed far faster than would be expected
by drift alone. This study demonstrates viral genetic turnover
within a focal population and the potential importance of adaptive
evolution in viral epidemic expansion (Bennett et al., 2003).
Finally, the Latin American isolates of DENV-3 genotype III ana-
lysed herein showed distinctive amino acid changes in the RNA-
dependent RNA polymerase (RdRp). It is known that all pathogenic
flaviviruses examined thus far inhibit host interferon responses. In
this regard, critical residues in the RdRp domain (355–735) have
been defined for the interferon antagonist function of Langat virus
NS5 protein (Park et al., 2007).
Long term and short-term studies reveal the association of viral
evolution with observed increased severity during epidemics. De-
spite some advances, the relevance of particular mutations has
not been investigated thoroughly. This is in large part due to the
lack of suitable animal models with which to study dengue virus
‘‘virulence’’. Variation in non-structural proteins has been associ-
ated with increasing severity in epidemiological settings corre-
sponding to endemic (Bennett et al., 2003;Chen et al., 2008;
Klungthong et al., 2004; Tang et al., 2010; Zhao et al., 2010) and
non-endemic dengue transmission (Rodriguez-Roche et al.,
2005a,b).
Moreover, evidences indicate that non-structural genetic ele-
ments may modulate reduced mosquito competence (Brault
et al., 2011).
New episodes of genotype replacement are frequently observed
(Vu et al., 2010). However, there are multiple factors implicated in
the transmission dynamic of DENVs that remain unclear (reviewed
by Holmes, 2009). Epidemiological data have suggested that fitness
is always context dependent and that as the immunological land-
scape changes, viral lineages that evade cross-immunity will be
at a selective advantage (Holmes, 2009). In addition, recent studies
suggest that strain diversity may limit the efficacy of monoclonal
antibody therapy or tetravalent vaccines against DENV as neutral-
isation potency generally correlated with a narrowed genotype
specificity (Brien et al., 2010).
5. Conclusions
Collectively, the results presented herein suggest that NS
proteins may play an important role in DENV virus evolution,
particularly NS1, NS2A and NS4B proteins, although experimental
confirmation based on a reverse genetics system are required.
The phylogenetic data provide evidence to suggest that multiple
introductions of DENV have occurred from the Latin American re-
gion into Venezuela and vice versa. The implications of significant
viral genetic diversity generated during hyperendemic transmis-
sion should be considered with regard to the development and
use of tetravalent vaccines. Particular changes in NS proteins could
modify viral phenotype, not only in terms of virulence; indeed the
identification of changes associated to potential naturally attenu-
ated strains would bring relevant knowledge to the rational design
of antivirals and dengue vaccine candidates.
Competing interests
The authors declare that they have no competing interests.
Acknowledgements
We thank Prof. Ernie A. Gould and Cameron P. Simons for rele-
vant suggestions and useful comments concerning the manuscript.
This research was supported by the European Commission
Grant No. 517708, 6th Framework Programme. The funders had
no role in study design, data collection and analysis, decision to
publish, or preparation of the manuscript.
References
Alexander, N., Balmaseda, A., Coelho, I.C., Dimaano, E., Hien, T.T., Hung, N.T., Janisch,
T., Kroeger, A., Lum, L.C., Martinez, E., Siqueira, J.B., Thuy, T.T., Villalobos, I.,
Villegas, E., Wills, B., (2011). Multicentre prospective study on dengue
classification in four South-east Asian and three Latin American countries.
Trop. Med. Int. Health doi:10.1111/j.1365-3156.2011.02793.x.
Alvarez, M., Rodriguez-Roche, R., Bernardo, L., Vazquez, S., Morier, L., Gonzalez, D.,
Castro, O., Kouri, G., Halstead, S.B., Guzman, M.G., 2006. Dengue hemorrhagic
fever caused by sequential dengue 1–3 virus infections over a long time
interval: Havana epidemic, 2001–2002. Am. J. Trop. Med. Hyg. 75 (6), 1113–
1117.
Bennett, S.N., Holmes, E.C., Chirivella, M., Rodriguez, D.M., Beltran, M., Vorndam, V.,
Gubler, D.J., McMillan, W.O., 2003. Selection-driven evolution of emergent
dengue virus. Mol. Biol. Evol. 20 (10), 1650–1658.
Betts, M.J., Russell, R.B., 2003. Amino acid properties and consequences of
substitutions. In: Barnes, M.R., Gray, I.C. (Eds.), Bioinformatics for Geneticists.
Wiley, New Jersey, pp. 289–316.
Brault, A.C., Kinney, R.M., Maharaj, P.D., Green, E.N., Reisen, W.K., Huang, C.Y., 2011.
Replication of the primary dog kidney-53 dengue 2 virus vaccine candidate in
Aedes aegypti is modulated by a mutation in the 5’ untranslated region and
amino acid substitutions in nonstructural proteins 1 and 3. Vector Borne
Zoonotic. Dis. 11 (6), 683–689.
Brien, J.D., Austin, S.K., Sukupolvi-Petty, S., O’Brien, K.M., Johnson, S., Fremont, D.H.,
Diamond, M.S., 2010. Genotype-specific neutralization and protection by
antibodies against dengue virus type 3. J. Virol. 84 (20), 10630–10643.
Chambers, T.J., McCourt, D.W., Rice, C.M., 1989. Yellow fever virus proteins NS2A,
NS2B, and NS4B: identification and partial N-terminal amino acid sequence
analysis. Virology 169 (1), 100–109.
Chen, H.L., Lin, S.R., Liu, H.F., King, C.C., Hsieh, S.C., Wang, W.K., 2008. Evolution of
dengue virus type 2 during two consecutive outbreaks with an increase in
severity in southern Taiwan in 2001-2002. Am. J. Trop. Med. Hyg. 79 (4), 495–
505.
Chen, M.C., Lin, C.F., Lei, H.Y., Lin, S.C., Liu, H.S., Yeh, T.M., Anderson, R., Lin, Y.S.,
2009. Deletion of the C-terminal region of dengue virus nonstructural protein 1
(NS1) abolishes anti-NS1-mediated platelet dysfunction and bleeding tendency.
J. Immunol. 183 (3), 1797–1803.
Christenbury, J.G., Aw, P.P., Ong, S.H., Schreiber, M.J., Chow, A., Gubler, D.J.,
Vasudevan, S.G., Ooi, E.E., Hibberd, M.L., 2010. A method for full genome
sequencing of all four serotypes of the dengue virus. J. Virol. Methods 169 (1),
202–206.
Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7, 214.
Garcia, G., Sierra, B., Perez, A.B., Aguirre, E., Rosado, I., Gonzalez, N., Izquierdo, A.,
Pupo, M., Danay Diaz, D.R., Sanchez, L., Marcheco, B., Hirayama, K., Guzman,
M.G., 2010. Asymptomatic Dengue infection in a Cuban population confirms the
protective role of the RR variant of the FcgammaRIIa polymorphism. Am. J. Trop.
Med. Hyg. 82 (6), 1153–1156.
Goncalvez, A.P., Escalante, A.A., Pujol, F.H., Ludert, J.E., Tovar, D., Salas, R.A., Liprandi,
F., 2002. Diversity and evolution of the envelope gene of dengue virus type 1.
Virology 303 (1), 110–119.
342 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344

Grant, D., Tan, G.K., Qing, M., Ng, J.K., Yip, A., Zou, G., Xie, X., Yuan, Z., Schreiber, M.J.,
Schul, W., Shi, P.Y., Alonso, S., 2011. A single amino acid in nonstructural protein
NS4B confers virulence to dengue virus in AG129 mice through enhancement of
viral RNA synthesis. J. Virol. 85 (15), 7775–7787.
Gubler, D.J., 1997. The emergence of dengue/dengue haemorrhagic fever as a global
public health problem. In: Saluzzo, J.F., Dodet, B. (Eds.), Factors in the
Emergence of Arbovirus Diseases. Elsevier, Paris, pp. 83–92.
Gubler, D.J., Meltzer, M., 1999. Impact of dengue/dengue hemorrhagic fever on the
developing world. Adv. Virus Res. 53, 35–70.
Guzman, M.G., 1995. Dengue type 3 infection-Nicaragua and Panama October–
November, 1994. MMWR 44, 21–24.
Guzman, M.G., 2005. Global voices of science. Deciphering dengue: the Cuban
experience. Science 309 (5740), 1495–1497.
Guzman, M.G., Deubel, V., Pelegrino, J.L., Rosario, D., Marrero, M., Sariol, C., Kouri, G.,
1995. Partial nucleotide and amino acid sequences of the envelope and the
envelope/nonstructural protein-1 gene junction of four dengue-2 virus strains
isolated during the 1981 Cuban epidemic. Am. J. Trop. Med. Hyg. 52 (3), 241–
246.
Guzman, M.G., Kouri, G., 2008. Dengue haemorrhagic fever integral hypothesis:
confirming observations, 1987–2007. Trans. R. Soc. Trop. Med. Hyg. 102 (6),
522–523.
Guzman, M.G., Kouri, G., Bravo, J., Valdes, L., Vazquez, S., Halstead, S.B., 2002a. Effect
of age on outcome of secondary dengue 2 infections. Int. J. Infect. Dis. 6 (2),
118–124.
Guzman, M.G., Kouri, G., Valdes, L., Bravo, J., Alvarez, M., Vazques, S., Delgado, I.,
Halstead, S.B., 2000. Epidemiologic studies on Dengue in Santiago de Cuba,
1997. Am. J. Epidemiol. 152 (9), 793–799.
Guzman, M.G., Kouri, G., Valdes, L., Bravo, J., Vazquez, S., Halstead, S.B., 2002b.
Enhanced severity of secondary dengue-2 infections: death rates in 1981 and
1997 Cuban outbreaks. Rev. Panam. Salud Publica 11 (4), 223–227.
Guzman, M.G., Vazquez, S., Martinez, E., Alvarez, M., Rodriguez, R., Kouri, G., De Los
Reyes, J., Acevedo, F., 1997. Dengue in Nicaragua, 1994: reintroduction of
serotype 3 in the Americas. Rev. Panam. Salud Publica 1 (3), 193–199.
Hanley, K.A., Manlucu, L.R., Gilmore, L.E., Blaney Jr., J.E., Hanson, C.T., Murphy, B.R.,
Whitehead, S.S., 2003. A trade-off in replication in mosquito versus mammalian
systems conferred by a point mutation in the NS4B protein of dengue virus type
4. Virology 312 (1), 222–232.
Henchal, E.A., McCown, J.M., Seguin, M.C., Gentry, M.K., Brandt, W.E., 1983. Rapid
identification of dengue virus isolates by using monoclonal antibodies in an
indirect immunofluorescence assay. Am. J. Trop. Med. Hyg. 32 (1), 164–169.
Holmes, E.C., 2009. RNA virus genomics: a world of possibilities. J. Clin. Invest. 119
(9), 2488–2495.
Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: bayesian inference of phylogenetic
trees. Bioinformatics 17 (8), 754–755.
Kelly, E.P., Puri, B., Sun, W., Falgout, B., 2009. Identification of mutations in a
candidate dengue 4 vaccine strain 341750 PDK20 and construction of a full-
length cDNA clone of the PDK20 vaccine candidate. Vaccine 28 (17), 3030–3037.
Klungthong, C., Zhang, C., Mammen, M.P. Jr., Ubol, S., Holmes, E.C., 2004. The
molecular epidemiology of dengue virus serotype 4 in Bangkok, Thailand.
Virology 329 (1), 168–179.
Kouri, G., Guzman, M.G., Bravo, J., 1986. Hemorrhagic dengue in Cuba: history of an
epidemic. Bull. Pan. Am. Health Organ. 20 (1), 24–30.
Kouri, G., Guzman, M.G., Valdes, L., Carbonel, I., del Rosario, D., Vazquez, S., Laferte,
J., Delgado, J., Cabrera, M.V., 1998. Reemergence of dengue in Cuba: a 1997
epidemic in Santiago de Cuba. Emerg. Infect. Dis. 4 (1), 89–92.
Kouri, G.P., Guzman, M.G., Bravo, J.R., 1987. Why dengue haemorrhagic fever in
Cuba? 2. An integral analysis. Trans. R. Soc. Trop. Med. Hyg. 81 (5), 821–823.
Kummerer, B.M., Rice, C.M., 2002. Mutations in the yellow fever virus nonstructural
protein NS2A selectively block production of infectious particles. J. Virol. 76
(10), 4773–4784.
Lanciotti, R.S., Calisher, C.H., Gubler, D.J., Chang, G.J., Vorndam, A.V., 1992.
Rapid detection and typing of dengue viruses from clinical samples by using
reverse transcriptase-polymerase chain reaction. J. Clin. Microbiol. 30 (3), 545–
551.
Lanciotti, R.S., Gubler, D.J., Trent, D.W., 1997. Molecular evolution and phylogeny of
dengue-4 viruses. J. Gen. Virol. 78 (Pt 9), 2279–2284.
Laue, T., Emmerich, P., Schmitz, H., 1999. Detection of dengue virus RNA in patients
after primary or secondary dengue infection by using the TaqMan automated
amplification system. J. Clin. Microbiol. 37 (8), 2543–2547.
Leung, J.Y., Pijlman, G.P., Kondratieva, N., Hyde, J., Mackenzie, J.M., Khromykh, A.A.,
2008. The role of nonstructural protein Ns2a in flavivirus assembly. J. Virol. 82
(10), 4731–4741.
Lindenbach, B.D., Rice, C.M., 1997. Trans-Complementation of yellow fever virus
NS1 reveals a role in early RNA replication. J. Virol. 71 (12), 9608–9617.
Liu, W.J., Chen, H.B., Khromykh, A.A., 2003. Molecular and functional analyses of
Kunjin virus infectious cDNA clones demonstrate the essential roles for NS2A in
virus assembly and for a nonconservative residue in NS3 in RNA replication. J.
Virol. 77 (14), 7804–7813.
Liu, W.J., Chen, H.B., Wang, X.J., Huang, H., Khromykh, A.A., 2004. Analysis of
adaptive mutations in Kunjin virus replicon RNA reveals a novel role for the
flavivirus nonstructural protein NS2A in inhibition of beta interferon promoter-
driven transcription. J. Virol. 78 (22), 12225–12235.
Liu, W.J., Wang, X.J., Clark, D.C., Lobigs, M., Hall, R.A., Khromykh, A.A., 2006. A single
amino acid substitution in the West Nile virus nonstructural protein NS2A
disables its ability to inhibit alpha/beta interferon induction and attenuates
virus virulence in mice. J. Virol. 80 (5), 2396–2404.
Liu, W.J., Wang, X.J., Mokhonov, V.V., Shi, P.Y., Randall, R., Khromykh, A.A., 2005.
Inhibition of interferon signaling by the New York 99 strain and Kunjin subtype
of West Nile virus involves blockage of STAT1 and STAT2 activation by
nonstructural proteins. J. Virol. 79 (3), 1934–1942.
Mackenzie, J.M., Jones, M.K., Young, P.R., 1996. Immunolocalization of the dengue
virus nonstructural glycoprotein NS1 suggests a role in viral RNA replication.
Virology 220 (1), 232–240.
Mackenzie, J.M., Khromykh, A.A., Jones, M.K., Westaway, E.G., 1998. Subcellular
localization and some biochemical properties of the flavivirus Kunjin
nonstructural proteins NS2A and NS4A. Virology 245 (2), 203–215.
Mathew, A., Rothman, A.L., 2008. Understanding the contribution of cellular
immunity to dengue disease pathogenesis. Immunol. Rev. 225, 300–313.
McElroy, K.L., Tsetsarkin, K.A., Vanlandingham, D.L., Higgs, S., 2006. Manipulation of
the yellow fever virus non-structural genes 2A and 4B and the 3’non-coding
region to evaluate genetic determinants of viral dissemination from the Aedes
aegypti midgut. Am. J. Trop. Med. Hyg. 75 (6), 1158–1164.
Munoz-Jordan, J.L., Sanchez-Burgos, G.G., Laurent-Rolle, M., Garcia-Sastre, A., 2003.
Inhibition of interferon signaling by dengue virus. Proc. Natl. Acad. Sci. USA 100
(24), 14333–14338.
Muylaert, I.R., Galler, R., Rice, C.M., 1997. Genetic analysis of the yellow fever virus
NS1 protein: identification of a temperature-sensitive mutation which blocks
RNA accumulation. J. Virol. 71 (1), 291–298.
PAHO (1979). Dengue in the Caribbean, 1977. Proceedings of a Workshop held in
Montego Bay, Jamaica, 8-11 May, 1978. PAHO Scientific Publication 375.
PAHO (1990). Dengue hemorrhagic fever in Venezuela. Epidemiol. Bull. 11, 7–9.
Park, G.S., Morris, K.L., Hallett, R.G., Bloom, M.E., Best, S.M., 2007. Identification of
residues critical for the interferon antagonist function of Langat virus NS5
reveals a role for the RNA-dependent RNA polymerase domain. J. Virol. 81 (13),
6936–6946.
Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution.
Bioinformatics 14 (9), 817–818.
Ramirez, A., Fajardo, A., Moros, Z., Gerder, M., Caraballo, G., Camacho, D., Comach, G.,
Alarcon, V., Zambrano, J., Hernandez, R., Moratorio, G., Cristina, J., Liprandi, F.,
2010. Evolution of dengue virus type 3 genotype III in Venezuela:
diversification, rates and population dynamics. Virol. J. 7, 329.
Rey, F.A., Heinz, F.X., Mandl, C., Kunz, C., Harrison, S.C., 1995. The envelope
glycoprotein from tick-borne encephalitis virus at 2 A resolution. Nature 375
(6529), 291–298.
Rodriguez-Roche, R., Alvarez, M., Gritsun, T., Halstead, S., Kouri, G., Gould, E.A.,
Guzman, M.G., 2005a. Virus evolution during a severe dengue epidemic in Cuba,
1997. Virology 334 (2), 154–159.
Rodriguez-Roche, R., Alvarez, M., Guzman, M.G., Morier, L., Kouri, G., 2000.
Comparison of rapid centrifugation assay with conventional tissue culture
method for isolation of dengue 2 virus in C6/36-HT cells. J. Clin. Microbiol. 38
(9), 3508–3510.
Rodriguez-Roche, R., Alvarez, M., Holmes, E.C., Bernardo, L., Kouri, G., Gould, E.A.,
Halstead, S., Guzman, M.G., 2005b. Dengue virus type 3, Cuba, 2000–2002.
Emerg. Infect. Dis. 11 (5), 773–774.
Rodriguez-Roche, R., Sanchez, L., Burgher, Y., Rosario, D., Alvarez, M., Kouri, G.,
Halstead, S.B., Gould, E.A., Guzman, M.G., 2011. Virus role during intraepidemic
increase in dengue disease severity. Vector Borne Zoonotic. Dis. 11, 675–681.
Salas, R.A., Tovar, D., Barreto, A., de Miller, E., Leitmeyer, K., Rico-Hesse, R., 1998.
Serotypes and genotypes of dengue virus circulating in Venezuela, 1990–1997.
Acta Cient. Venez. 49 (Suppl. 1), 33–37.
San Martin, J.L., Brathwaite, O., Zambrano, B., Solorzano, J.O., Bouckenooghe, A.,
Dayan, G.H., Guzman, M.G., 2010. The epidemiology of dengue in the Americas
over the last three decades: a worrisome reality. Am. J. Trop. Med. Hyg. 82 (1),
128–135.
Sangkawibha, N., Rojanasuphot, S., Ahandrik, S., Viriyapongse, S., Jatanasen, S.,
Salitul, V., Phanthumachinda, B., Halstead, S.B., 1984. Risk factors in dengue
shock syndrome: a prospective epidemiologic study in Rayong, Thailand. I. The
1980 outbreak. Am. J. Epidemiol. 120 (5), 653–669.
Schmidt, D.J., Pickett, B.E., Camacho, D., Comach, G., Xhaja, K., Lennon, N.J., Rizzolo,
K., de Bosch, N., Becerra, A., Nogueira, M.L., Mondini, A., da Silva, E.V.,
Vasconcelos, P.F., Munoz-Jordan, J.L., Santiago, G.A., Ocazionez, R., Gehrke, L.,
Lefkowitz, E.J., Birren, B.W., Henn, M.R., Bosch, I., (2011). A phylogenetic analysis
using full-length viral genomes of South American dengue serotype 3 in
consecutive Venezuelan outbreaks reveals a novel NS5 mutation. Infect Genet
Evol. 11 (8), 2011–2019.
Schreiber, M.J., Holmes, E.C., Ong, S.H., Soh, H.S., Liu, W., Tanner, L., Aw, P.P., Tan,
H.C., Ng, L.C., Leo, Y.S., Low, J.G., Ong, A., Ooi, E.E., Vasudevan, S.G., Hibberd, M.L.,
2009. Genomic epidemiology of a dengue virus epidemic in urban Singapore. J.
Virol. 83 (9), 4163–4173.
Seijo, A., 2009. Dengue 2009: chronology of an epidemic. Arch. Argent. Pediatr. 107
(5), 387–389.
Sierra, B., Alegre, R., Perez, A.B., Garcia, G., Sturn-Ramirez, K., Obasanjo, O., Aguirre,
E., Alvarez, M., Rodriguez-Roche, R., Valdes, L., Kanki, P., Guzman, M.G., 2007.
HLA-A, -B, -C, and -DRB1 allele frequencies in Cuban individuals with
antecedents of dengue 2 disease: Advantages of the Cuban population for
HLA studies of dengue virus infection. Hum. Immunol. 68 (6), 531–540.
Sierra, B.D.L.C., Garcia, G., Perez, A.B., Morier, L., Alvarez, M., Kouri, G., and Guzman,
M. G. (2006). Ethnicity and difference in dengue virus-specific memory T cell
responses in Cuban individuals. Viral Immunol. 19(4), 662–668.
Steel, A., Gubler, D.J., Bennett, S.N., 2010. Natural attenuation of dengue virus type-2
after a series of island outbreaks: a retrospective phylogenetic study of events
in the South Pacific three decades ago. Virology 405 (2), 505–512.
R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344 343

Swofford, D.L., 2003. PAUP⁄: Phylogenetic Analysis Using Parsimony (⁄and Other
Methods). Sinauer Associates, Sunderland, Massachusetts.
Tajima, S., Takasaki, T., Kurane, I., 2011. Restoration of replication-defective dengue
type 1 virus bearing mutations in the N-terminal cytoplasmic portion of NS4A
by additional mutations in NS4B. Arch. Virol. 156 (1), 63–69.
Tan, G.K., Alonso, S., 2009. Pathogenesis and prevention of dengue virus infection:
state-of-the-art. Curr. Opin. Infect. Dis. 22 (3), 302–308.
Tang, Y., Rodpradit, P., Chinnawirotpisan, P., Mammen, M.P. Jr., Li, T., Lynch, J.A.,
Putnak, R., Zhang, C., 2010. Comparative analysis of full-length genomic
sequences of 10 dengue serotype 1 viruses associated with different
genotypes, epidemics, and disease severity isolated in Thailand over 22 years.
Am. J. Trop. Med. Hyg. 83 (5), 1156–1165.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The
ClustalX windows interface. Flexible strategies for multiple sequence alignment
aided by quality analysis tools. Nucleic Acids Res. 24, 4876–4882.
Twiddy, S.S., Farrar, J.J., Vinh Chau, N., Wills, B., Gould, E.A., Gritsun, T.,
Lloyd, G., Holmes, E.C., 2002. Phylogenetic relationships and differential
selection pressures among genotypes of dengue-2 virus. Virology 298 (1), 63–
72.
Uzcategui, N.Y., Camacho, D., Comach, G., Cuello de Uzcategui, R., Holmes, E.C.,
Gould, E.A., 2001. Molecular epidemiology of dengue type 2 virus in Venezuela:
evidence for in situ virus evolution and recombination. J. Gen. Virol. 82, 2945–
2953.
Uzcategui, N.Y., Comach, G., Camacho, D., Salcedo, M., Cabello de Quintana, M.,
Jimenez, M., Sierra, G., Cuello de Uzcategui, R., James, W.S., Turner, S., Holmes,
E.C., Gould, E.A., . Molecular epidemiology of dengue virus type 3 in Venezuela.
J. Gen. Virol. 84 (Pt 6), 1569–1575.
Vaughn, D.W., Green, S., Kalayanarooj, S., Innis, B.L., Nimmannitya, S., Suntayakorn,
S., Endy, T.P., Raengsakulrach, B., Rothman, A.L., Ennis, F.A., Nisalak, A., 2000.
Dengue viremia titer, antibody response pattern, and virus serotype correlate
with disease severity. J. Infect. Dis. 181 (1), 2–9.
Vu, T.T., Holmes, E.C., Duong, V., Nguyen, T.Q., Tran, T.H., Quail, M., Churcher, C.,
Parkhill, J., Cardosa, J., Farrar, J., Wills, B., Lennon, N.J., Birren, B.W., Buchy, P.,
Henn, M.R., Simmons, C.P., 2010. Emergence of the Asian 1 genotype of dengue
virus serotype 2 in Vietnam: in vivo fitness advantage and lineage replacement
in South-East Asia. PLoS Negl. Trop. Dis. 4 (7), e757.
Wallis, T.P., Huang, C.Y., Nimkar, S.B., Young, P.R., Gorman, J.J., 2004. Determination
of the disulfide bond arrangement of dengue virus NS1 protein. J Biol Chem. 279
(20), 20729–20741.
Weaver, S.C., Vasilakis, N., 2009. Molecular evolution of dengue viruses:
contributions of phylogenetics to understanding the history and epidemiology
of the preeminent arboviral disease. Infect. Genet. Evol. 9 (4), 523–540.
WHO (1997). Dengue haemorrhagic fever: diagnosis, treatment, prevention and
control. 2nd ed. WHO, Geneva.
WHO/TDR (2009). Dengue guidelines for diagnosis, treatment, prevention and
control. New Edition. WHO Press, Geneva.
Zhao, R., Chinnawirotpisan, P., Klungthong, C., Zhang, C., Putnak, R., 2010. Evidence
for inter- and intra-genotypic variations in dengue serotype 4 viruses
representing predominant and non-predominant genotypes co-circulating in
Thailand from 1977 to 2001. Virus Genes 41 (1), 5–13.
344 R. Rodriguez-Roche et al. / Infection, Genetics and Evolution 12 (2012) 332–344