Elsevier Editorial System(tm) for Infection, Genetics and Evolution Manuscript Draft Manuscript Number: MEEGID-D-14-00393R1 Title: Phylogeny of Dengue and Chikungunya viruses in Al Hodayda governorate, Yemen. Article Type: Research paper Keywords: DENV, CHIKV, phylogeny Corresponding Author: Dr. massimo ciccozzi, Corresponding Author's Institution: National Institute of Health First Author: massimo ciccozzi Order of Authors: massimo ciccozzi; Alessandra Lo Presti; Eleonora Cella; Marta Giovanetti; Alessia Lai; Gamal El-Sawaf; Giovanni Faggioni; Fenicia Vescio; Ranya Al Ameri; Riccardo De Santis; Ghada Helaly; Alice Pomponi; Dalia Metwally; Massimo Fantini; Hussein Qadi; Gianguglielmo Zehender; Florigio Lista; Giovanni Rezza Abstract: Yemen, which is located in the southwestern end of the Arabian Peninsula, is one of countries most affected by recurrent epidemics caused by emerging vector-borne viruses. Dengue virus (DENV) outbreaks have been reported with increasing frequency in several governorates since the year 2000, and the chikungunya virus (CHIKV) has been also responsible of large outbreaks and it is now a major public health problem in Yemen. We report the results of the phylogenetic analysis of DENV-2 and CHIKV isolates (NS1 and E1 genes, respectively) detected in an outbreak occurred in al-Hodayda in 2012. Estimates of the introduction date of CHIKV and DENV-2, and the phylogeographic analysis of DENV-2 are also presented. Phylogenetic analysis showed that the Yemen isolates of DENV belonged to the lineage 2 Cosmopolitan subtype, whereas CHIKV isolates from Yemen belonged to the ECSA genotype. All the CHIKV isolates from Yemen were statistically supported and dated back to the year 2010 (95% HPD: 2009-2011); these sequences showed an alanine in the aminoacid position 226 of the E1 protein. Phylogeographic analysis of DENV-2 virus showed that cluster 1, which included Yemen isolates, dated back to 2003 Burkina Faso strains (95%HPD 1999-2007). The Yemen, cluster dated back to 2011 (95% HPD 2009-2012). Our study sheds light on the global spatiotemporal dynamics of DENV-2 and CHIKV in Yemen. This study reinforces both the need to monitor the spread of CHIKV and DENV, and to apply significant measures for vector control.

Michel Tibayrenc

UM1-CNRS 5290-IRD 224, MIVEGEC/IDVEGEC

(Infectious Diseases and Vectors, Ecology, Genetics, Evolution and Control),

Centre National de la Recherche Scientifique (CNRS), IRD Center,

BP 64501, 34394 Montpellier Cedex 5, France,

Email: michel.tibayrenc@ird.fr

Dear Editor

We submitted the revised version of the paper entitled “Phylogeny of Dengue and Chikungunya

viruses in Al Hodayda governorate, Yemen” . We answered point by point all the questions by

the reviewers. All authors have seen and approved this version of the manuscript.

Thank you for considering this paper for publication in your journal.

Sincerely,

Massimo Ciccozzi

Department of Infectious Parasitic and Immunomediated Diseases, Reference Centre on Phylogeny,

Molecular Epidemiology and Microbial Evolution (FEMEM)/ Epidemiology Unit, National

Institute of Health, Rome, Italy

e-mail: ciccozzi@iss.it; Phone number: +390649903187

Cover Letter

Reviewers' comments:

Reviewer #1: The authors have reported a phylogenetic analysis of dengue and chikungunya viruses

recovered in Yemen. This is ,essentially, a descriptive report that does not provide any surprises and

does not appear to provide any insights into the how and why of global movement of arboviruses

and their establishment of transmission cycles at new localities. The analyses appear to have been

undertaken with a rigor beyond that required for a study of this kind e.g. Fig S1 - but for which the

authors might be commended.

Q: The Reviewer could find no rationale for the analysis of dengue NS1 genes rather than E genes,

particularly as the database of E gene sequences is much larger than that of NS1 sequences.

A: We agree in part with the Reviewer about the database. Indeed NS1 was used for DENV and

E gene for CHIKV and both the dataset were built used specific criteria as reported in material

and methods.

Q:The Abstract could be shorter and the question being asked and the significance of the

conclusion(s) made clearer.

A: As suggested by the Review we modified the abstract.

Q: The manuscript was easy to read and to comprehend although there are some grammatical issues

that should be addressed e.g. the appropriate use of "the". However, the manuscript is too long for

the information conveyed.

A:Done.

Q:The authors should consult the Guide to Authors for the correct use of capitals when writing the

name of a virus. Dengue is a disease not a virus e.g. a mosquito transmits dengue viruses, not

dengue.

A:Thanks to the Review. We have corrected as requested.

Q:The figures are clear but there are more figures in the text than are necessary. A phylogenetic tree

for DENV sequences and another for CHIKV sequences would have been adequate.

Some of the data relating to times to recent ancestors have probably been over-interpreted.

A:We disagree with the Review, alla the figures in the manuscript are important to describe the

outbreaks of these two viruses, indeed we put as supplementary figures the only two not full

necessary.

Reviewer #2: The manuscript described the genetic analysis of Dengue and Chikungunya viruses in

Al Hodayda governorate, Yemen. Even the data are not new, information of these viruses in Yemen

is limited. Therefore, it is worthy to publish it in the journal.

Two comments:

Q: 1. English has some grammatical error. The manuscript should be checked by a native English

speaker.

A:Done.

Q: 2. Some parts of data were published in EID journal. Authors did cite this paper in the

manuscript. A complete citation such as title, year... is needed.

A:Done.

*Reply to Reviewers

Highlights

Phylogenetic analysis of DENV-2 and CHIKV isolates detected in Yemen was reported.

Estimates of introduction date of CHIKV and DENV-2 was presented.

Phylogeographic analysis of DENV-2 is presented.

This study reinforces the need to monitor the spread of CHIKV and DENV

*Highlights (for review)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1

Phylogeny of Dengue and Chikungunya viruses in Al Hodayda governorate, Yemen.

Massimo Ciccozzi1,2*

, Alessandra Lo Presti1, Eleonora Cella

1, Marta Giovanetti

1, Alessia Lai

3,

Gamal El-Sawaf4, Giovanni Faggioni

5, Fenicia Vescio

1, Ranya Al Ameri

6, Riccardo De Santis

5,

Ghada Helaly4, Alice Pomponi

5, Dalia Metwally

4, Massimo Fantini

7, Hussein Qadi

6, Gianguglielmo

Zehender3, Florigio Lista

5, Giovanni Rezza

1.

1Department of Infectious, Parasitic and Immunomediated Diseases, Istituto Superiore di Sanità,

Rome, Italy

2University Hospital Campus Bio-Medico

3Department of Biomedical and Clinical Science, Infectious Diseases and Immunopathology

Section, „L. Sacco‟ Hospital, University of Milan, Milan, Italy

4Medical Research Institute, Alexandria University, Egypt

5Histology and Molecular Biology Section, Army Medical and Veterinary Research Center, Rome,

Italy

6University of Sana‟a, Republic of Yemen

7Department of Clinical Sciences and Translational Medicine, University of Rome “Tor Vergata”,

Rome, Italy

*Corresponding author: Massimo Ciccozzi

Department of Infectious Parasitic and Immunomediated Diseases, Istituto Superiore di Sanità,

Rome, Italy

e-mail: ciccozzi@iss.it ; Phone number: +390649903187

*ManuscriptClick here to view linked References

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

2

ABSTRACT

Yemen, which is located in the southwestern end of the Arabian Peninsula, is one of countries most

affected by recurrent epidemics caused by emerging vector-borne viruses. Dengue virus (DENV)

outbreaks have been reported with increasing frequency in several governorates since the year 2000,

and the chikungunya virus (CHIKV) has been also responsible of large outbreaks and it is now a

major public health problem in Yemen. We report the results of the phylogenetic analysis of

DENV-2 and CHIKV isolates (NS1 and E1 genes, respectively) detected in an outbreak occurred in

al-Hodayda in 2012. Estimates of the introduction date of CHIKV and DENV-2, and the

phylogeographic analysis of DENV-2 are also presented. Phylogenetic analysis showed that the

Yemen isolates of DENV belonged to the lineage 2 Cosmopolitan subtype, whereas CHIKV

isolates from Yemen belonged to the ECSA genotype. All the CHIKV isolates from Yemen were

statistically supported and dated back to the year 2010 (95% HPD: 2009-2011); these sequences

showed an alanine in the aminoacid position 226 of the E1 protein. Phylogeographic analysis of

DENV-2 virus showed that cluster 1, which included Yemen isolates, dated back to 2003 Burkina

Faso strains (95%HPD 1999-2007). The Yemen, cluster dated back to 2011 (95% HPD 2009-2012).

Our study sheds light on the global spatiotemporal dynamics of DENV-2 and CHIKV in Yemen.

This study reinforces both the need to monitor the spread of CHIKV and DENV, and to apply

significant measures for vector control.

Keywords: DENV, CHIKV, phylogeny

1. Introduction

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

3

Arbovirus related infections are not uncommon in the Arabian peninsula as well as in other

countries in the Middle East and North Africa (Hotez et al., 2012). Dengue (DEN) outbreaks have

been reported in Djibouti and in the Arabian peninsula since 1990 (Ooi and Gubler, 2011), and the

Dengue virus (DENV) is now endemic in western and southern regions of Saudi Arabia (Ahmed,

2010).

Yemen, which is located in the southwestern end of the Arabian Peninsula, is one of countries most

affected by recurrent epidemics caused by emerging vector borne viruses. Outbreaks of DEN fever

have been reported with increasing frequency since the year 2000 in several governorates, and

Chikungunya virus (CHIKV) has been also responsible of large outbreaks (EpiSouth, 2011) and it is

now a major public health problem in Yemen (Zayed et al. 2012).

There is some evidence that several DENV serotypes circulated in Yemen over the last decades.

DENV-3 has been the predominant serotype for a long time. In fact, DENV-3 caused a major

outbreak in the western governorate of al-Hudayda in 2005 (WHO, 2009). The same serotype was

detected in travelers from Yemen (Ravanini et al., 2010) and during an outbreak in the southern

district of Hadramouuth (Ghouth et al., 2012). However, the identification of hemorrhagic

manifestations in patients infected with the DENV-3 genotype in Al Mukalla, the capital city of

Hadramouth governorate, suggested previous circulation of other DENV genotypes (Madani et al.,

2013). Moreover, high antibody titers against all four DENV serotypes, consistent with a secondary

heterotypic infection, in a Yemenite traveler affected by hemorrhagic DEN in 1983, may also

suggest that multiple introductions with further circulation of different DENV serotypes occurred

even in past decades (Jimenez-Lucho et al., 1984).

A study conducted in the governorate of Al-Hodayda in 2011 which involved 47 patients with

Dengue-like illness, detected DENV-1 in two cases and DENV-3 in one case, while no DENV-2

positive samples were identified (Rezza et al., 2014).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

4

During a recent outbreak of DENV-like illness in the same region in 2012, we found co-circulation

of several DENV and CHIKV genotypes (Rezza et al., 2014). DENV-2 was largely predominant

(41 cases out of 55 tested DENV RNA positive), but DENV-1 was also found in 2 cases. Thus,

whether DENV 2 serotype was newly introduced in Yemen in 2012 remained undefined.

Hereby we report the results of the phylogenetic analysis of DENV-2 and CHIKV isolates detected

in the al-Hodayda outbreak. Estimates of the introduction date of CHIKV and DENV-2, and the

phylogeographic analysis of DENV-2 is also presented.

2. Materials and methods

2.1 Patients, serum samples and genomic RNA

Four-hundred patients were consecutive enrolled by five hospital centers (Renal Center,

Maritime College, Al Rasheed, AI-Thawra, Al Salakhana) located in AI-Hodayda, Yemen

between January 2011 and June 2012. The case definition was fever (>37.5°C) and at least two

of the following symptoms: headache, joint pain, muscle pain, skin rash. Serum samples were

collected within four days from the date of hospital admission, and stored and shipped at -

20°C. Of the study participants, fifty-five were PCR positive for DENV RNA of which twelve

were untyped, forty-one were positive for DENV 2 and two were positive for DENV 1. Eleven

sera were PCR positive for CHIKV RNA. The study has been approved by local ethical

committee.

2.2 Analysis of nucleic acids

Samples tested positive with high viral RNA titer were chosen for sequencing analysis. Nine sera

which tested positive for DENV serotype 2 and eight sera which tested positive for CHIKV

underwent amplification and sequencing reactions for NS1 and E1 genes, respectively. The

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

5

amplification of DENV NS1 gene serotype 2, was performed using the oligonucleotide 5'-

GATAGTGGTTGCGTTGTGAGC-3' (forward) and 5'-AGCTGTGACCAAGGAGTTGAC-3'

(reverse). The CHIKV E1 gene, was amplified using the oligonucleotides 5'-

AGCGAAGCACATGTGGAGAAG-3' (forward) and 5-TTAGTGCCTGCTGAACGACAC-3

(reverse) . All the amplifications were performed by using the Superscript III One step RT-PCR kit

(Invitrogen, CA). The reaction conditions were 50°C x 30 minutes, 95°C x 2 minutes followed by

10 cycles of 95°C x 10 seconds, 54°C x 30 (T - 0.5 °C each cycle) seconds, 68°C x 1 minute, then

25 cycles of 95°C x 10 seconds, 54°C x 30 seconds, 68°C x 1 minute. The PCR products were

visualized on 1% agarose gel and purified by spin column (Macherey Nagel, Germany). Sequencing

reactions were carried out on the Ceq 8000 automated DNA sequencer (CEQ 8000, Beckman

Coulter), according to manufacturer‟s instructions. The obtained sequences for CHIKV and Dengue

viruses have been deposited in GenBank under the following accession numbers: from KJ742803 to

KJ742819.

2.3 Viruses dataset

Two dataset for DENV were built. The first one included 9 Dengue virus NS1 gene isolates plus 49

reference sequences downloaded from GenBank (http://www.ncbi.nlm.nih.gov/), which was used to

perform the lineage and subtype assignment. The second dataset included 9 Dengue 2 virus NS1

gene isolates previously classified as subtype Cosmopolitan plus 72 Dengue 2 virus NS1 gene

Cosmopolitan subtype sequences downloaded from NCBI (http://www.ncbi.nlm.nih.gov/). The

following inclusion criteria are used for sequences: (1) sequences already published in peer-

reviewed journals (2) no uncertainty about the subtype assignment; (3) known sampling date and

location. This dataset was built to estimate of the Dengue 2 virus Cosmopolitan subtype NS1 gene

mean evolutionary rate and to perform the phylogeography.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

6

Two different datasets for CHIKV were built. The first one included 8 CHIKV partial E1 gene

isolates from Yemen and 111 reference sequences downloaded from GenBank

(http://www.ncbi.nlm.nih.gov/genbank/) and was used to assign the genotype.

The second dataset included 126 CHIKV partial E1 sequences and was used to co-estimate the

evolutionary rate and time-scaled phylogeny. Collection dates of the sequences ranged from 1953 to

2012. Starting from the Bayesian tree of the second dataset, a clade including 70 sequences (8

isolates from Yemen, 6 from Buthan, 18 from India, 1 from France that was a case imported from

India, 3 from Singapore, 7 from Sri Lanka; 1 from Japan that was a Japanese returning from Sri

Lanka, 2 from Italy, 1 from Malaysia, 1 from Bangladesh, 4 from Thailand, 8 from Cambodia, 2

from Kenya, 2 from Comoros, 1 from Seychelles, 1 from Mauritius, 1 from Reunion, 3 from

China), was highlighted and used to study both the phylogenetic relationships between Yemen and

other countries and to date the epidemic. Moreover, the presence of the mutation A226V was

investigated.

The sequences were chosen on the basis of the same criteria described above for DENV.

The sequences of all the datasets were aligned using Clustal X (Ciccozzi et al., 2013) and manually

edited by Bioedit software v. 7.2.5. Modeltest 3.7 was used (Ciccozzi et al., 2013) for all the

datasets to select the simplest evolutionary model that adequately fitted the sequence data.

2.4 Likelihood mapping

The phylogenetic signal of each dataset was investigated by means of the likelihood mapping

analysis of 10,000 random quartets generated using TreePuzzle as already described (Zehender et

al., 2011; Ciccozzi et al, 2013). Groups of four randomly chosen sequences (quartets) were

evaluated. For each quartet the three possible unrooted trees were reconstructed using the maximum

likelihood approach under the selected substitution model. The posterior probabilities of each tree

were then plotted on a triangular surface so that fully resolved trees fall into the corners and the

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

7

unresolved quartets in the centre of the triangle (a star-tree). When using this strategy, if more than

30% of the dots fall into the centre of the triangle, the data are considered unreliable for the

purposes of phylogenetic inference.

2.5 Phylogenetic analysis

Maximum likelihood (ML) phylogenetic trees were estimated, on the first dataset for both Dengue

and CHIKV viruses, using the best fitting substitution model selected with ModelTest (Ciccozzi et

al., 2013), (TrN + I + G for Dengue virus and HKY + I+ G for CHIKV virus) with Phyml 3.0

(http://www.atgc-montpellier.fr/phyml/). The statistical robustness and reliability of the branching

order within the phylogenetic tree was confirmed with the bootstrap analysis (bootstrap value

≥70%).

2.6 Evolutionary rate estimate

The evolutionary rate was estimated by using a Bayesian MCMC approach (Beast v. 1.7.4,

http://beast.bio.ed.ac.uk) (Drummond et al., 2005; Drummond and Rambaut, 2007) implementing

the model selected with ModelTest using both a strict and an uncorrelated log-normal relaxed clock

model. The data set used were the second ones for both Dengue and Chikungunya viruses.

As coalescent priors, three parametric demographic models of population growth (constant size,

exponential, expansion growth) and a Bayesian skyline plot (BSP, a non-parametric piecewise-

constant model) were compared. The best fitting models were selected using a Bayes Factor (BF

with marginal likelihoods). In accordance with Kass and Raftery (Kass and Raftery, 1995), the

strength of the evidence against H0 was evaluated as follows: 2 ln BF < 2, no evidence; 2–6, weak

evidence; 6–10, strong evidence; > 10, very strong evidence. A negative value indicates evidence in

favor of H0. Only values of ≥ 6 were considered significant.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

8

Chains were conducted for at least 50 x 106

generations, and sampled every 5000 steps and for 150

106

generations, sampling every 15,000th generation for DENV and CHIKV respectively.

Convergence of the MCMC was assessed by calculating the ESS for each parameter. Only

parameter estimates with ESS‟s of > 250 were accepted. Maximum clade credibility trees were

obtained from the trees posterior distributions with the Tree-Annotator software v 1.7.4, included in

the Beast package (Drummond et al., 2005; Drummond and Rambaut, 2007). Statistical support for

specific monophyletic clades was assessed by calculating the posterior probability.

2.7 Phylogeographic analysis

The continuous-time Markov Chain (MCC) process over discrete sampling locations implemented

in BEAST (Drummond and Rambaut, 2007) was used for the geographical analysis, implementing

the Bayesian Stochastic Search Variable Selection (BSSVS) model which allows the diffusion rates

to be zero with a positive prior probability. Comparison of the posterior and prior probabilities of

the individual rates being zero provided a formal BF for testing the significance of the linkage

between locations. This analysis was performed only for the Dengue 2 virus NS1 region (second

dataset).

The sampling locations were, Yemen, Australia, Brunei Darussalam, Burkina Faso, China, East

Timor, Guam, Indonesia, India, Pakistan, Singapore, Sri Lanka, and Vietnam.

The final trees were manipulated in FigTree v. 1.4 for display. The most probable location of each

node was highlighted by labeling the branches with different colors.

3. Results

3.1 Likelihood mapping analysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

9

The phylogenetic noise of each data set was investigated by means of likelihood mapping. For the

Dengue virus polyprotein the percentage of dots falling in the central area of the triangles was 1.2%

and 2.0% for the first and second dataset, respectively (Fig. S1); since none of the dataset showed

more that 30% of noise, all of them contained sufficient phylogenetic signal. For the chikungunya

virus in each data set the percentage of dots falling in the central area of the triangles was 14.8 %

for the first dataset and 18.3 % for the second dataset (Fig. S2). As none of the datasets showed

more than 30% of noise, all of them showed a sufficient phylogenetic signal.

3.2 Phylogenetic analysis

The ML tree for DENV, showed distinct clusters of the different lineages and subtypes. The Yemen

isolates belonged to the lineage 2 Cosmopolitan subtype (Fig. 1). All the isolates clustered in the

same statistically supported clade, indicating a unique epidemic. Maximum likelihood (ML)

phylogenetic analysis for Chikungunya virus (Fig. 2) showed three distinct genotypes: West Africa,

Asia and East-central– South – African (ECSA) genotype. All the eight sequences from Yemen

belonged to the ECSA genotype and clustered inside the clade which include also the viruses

isolated during the Indian Ocean outbreak.

3.3 Evolutionary rate estimate

Bayes Factor analysis for Dengue virus showed that the relaxed clock fitted the data significantly

better than the strict clock (2lnBF between the strict and relaxed clock was 15 in favor of the

second). Under the relaxed clock, the BF analysis showed that the BSP was better than the other

models (2lnBF >24). The estimated mean value of DENV-2 Cosmopolitan subtype NS1 gene

evolutionary rate was 8.23 x 10-4

sub/site/year (95%HPD 6.62 x10-4

-1.01 x10-3

).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

10

Bayes Factor analysis for CHIKV, showed that the relaxed clock fitted the data significantly better

than the strict clock (2lnBF between the strict and relaxed clock was 90 in favor of the second).

Under the relaxed clock, the BF analysis showed that the BSP was better than the other models

(2lnBF > 100). The estimated mean value of the CHIKV E1 evolutionary rate, was 1.037 x 10-3

substitution/site/year (95% highest posterior density interval HPD: 5.15 x 10-4

- 1.716 x 10-3

).

Fig. 3 highlighted the Bayesian maximum clade credibility tree of the 70 CHIKV E1 sequences

which included the eight isolates from Yemen. Clade I was statistically supported and included five

sequences from India, the eight isolates from Yemen, a sequence from France that was a case

imported from India (from Rajasthan - Acc. Number FR846304) and six sequences from Buthan.

All the sequences from Yemen were included in a statistically supported cluster and dated back to

the year 2010 (95% HPD: 2009-2011).

3.4 Phylogeographic analysis of DENV-2 virus

Fig. 4 shows the time and location-annotated Bayesian maximum clade credibility tree of the

DENV-2 Cosmopolitan subtype NS1 gene sequences. The most probable locations of the internal

nodes are indicated by different colors, and the time to most recent common ancestor (tMRCA)

estimates under a relaxed molecular clock model are highlighted. The phylogeographic analysis

showed that the root of the tree dated back 1964 (HPD 95% 1939-1984) and highlighted three

distinct clades (A, B and C). Clade A originated in Indonesia and included sequences from

Australia, Brunei Darussalam, China, East Timor, Guam, Indonesia, Singapore, Vietnam. Clade B

originated in India and included sequences from China, India, Pakistan and Sri Lanka. Clade C

originated in Indonesia and included sequences from Burkina Faso, Indonesia and Yemen. All

these Clade has been statistically supported by a state probability ≥ 0.35.

Phylogeographic reconstruction indicated the most probable location of cluster 1 within Clade C,

which included the Yemen isolates and Burkina Faso sequences, probably was Burkina Faso and

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

11

dated back to 2003 (95%HPD 1999-2007). Moreover the Yemen cluster alone dated back to 2011

(95% HPD 2009-2012).

3.5 CHIKV A226V mutation analysis

The presence of the mutation A226V in the 70 CHIKV E1 sequences (which also included the eight

isolates from Yemen) reported in Figure 3 was investigated. The valine in the aminoacid position

226 of the E1 sequences was identified in the cluster which included the sequences from Reunion

and Mauritius and in the clade which included the sequences from India (isolated in 2007), Italy

(collected in 2007), Sri Lanka (collected in 2008), Malaysia (collected in 2008), Bangladesh

(collected in 2008), Singapore (collected in 2008), Thailand (collected in 2008- 2009), China

(collected in 2010), Cambodia (collected in 2011). All the other sequences did not showed the

mutation A226V. Thence, the CHIKV strains which circulated in Yemen showed an alanine in the

aminoacid position 226 of the E1 protein.

4. Discussion

We recently described a simultaneous outbreak of DENV and CHIKV in Al-Hodayda, a port city

located in Western Yemen, in 2012. Although there was some evidence that DENV-1 serotype, but

not DENV-2, was already circulating in 2011, we found that DENV-2 largely predominated in 2012

(Rezza et al., 2014). This finding suggested that the DENV-2 serotype could have been newly

introduced in the area of Al-Hodayda, causing to an explosive outbreak. Actually, our evolutionary

analysis confirmed that DENV-2 was introduced late in 2011, which is consistent with molecular

epidemiology data. With regard to CHIKV, we estimated that its introduction dates back to 2010,

which is rather consistent with the reports of a massive epidemic that occurred in the area of Al-

Hodayda in 2011 (EpiSouth, 2011). In particular, it should be mentioned that, in the same year,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

12

CHIKV was detected by PCR in pools of Aedes aegypti mosquitoes collected in the governorate of

Al-Hodayda (Zayed et al., 2012).

These findings suggest that vector-borne viruses, such as DENV and CHIKV, are continuously

reintroduced in Yemen. Specifically, the site of our study, Al Hodayda, is a port city located along

the Red Sea coast, in the so-called Tihamah (“hot lands” or “hot earth”). Red sea ports, with

movement of refugees and intense human activities, are the entry gates for infected humans and

vectors, representing also a favorable habitat for mosquitoes. Furthermore, water scarcity and lack

of infrastructure in peri-urban areas require storage of water for household, and water containers

also favor the breeding habitat of mosquitoes (Rezza et al., 2014). To this regard, a case-control

study conducted in Jeddah in 2004 identified a series of local and individual factors, such as the

presence of stagnant water in indoor drainage holes, indoor larvae near construction sites, and older

age, associated with mosquito-borne infections. On the other hand, face-to-face health education

significantly decreased the risk of infection (Kholedi et al., 2012).

Our phylogeographic analyses, based on DENV-2 Cosmopolitan subtype NS1 gene sequences

sampled from different countries worldwide, suggested that Burkina Faso acted as viral source

populations for the Yemen epidemic strain. This could be explained by the importation of travel-

associated Dengue cases from Africa (Jimenez-Lucho et al., 1984; Ravanini et al., 2010; Nicoletti et

al., 2008). Although the highest similarity was found with Burkina Faso strains, an alternative

hypothesis is that Yemen strains originated from India or Indonesia. The interpretation of this

analysis needs caution, since it relies on a limited number of data included in the web and then

biased by the lack of information on many epidemics occurred in tropical ares of Africa and Asia.

With regard to CHIKV virus, Yemen strains were similar to ECSA strains causing the Indian Ocean

outbreak after 2004. The lack of the A226V mutation is consistent with the predominance of Aedes

aegypti as CHIKV virus vector in Yemen (Zayed et al., 2012).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

13

In conclusion, the geographic spread of the DENV-2 virus strains and the dates of introduction (the

time-scaled phylogeny) of DENV-2 and CHIKV viruses in Yemen were estimated thorough a

sophisticated Bayesian statistical inference framework. Further understanding of the key source

areas of viral dispersal and the factors underlying the emergence and spread of fast evolving viruses

can help directing epidemiological surveillance efforts. This study reinforces the need to monitor

the spread of CHIKV and DENV viruses, and to reinforce vector control measures. This is a critical

area for future research that may ultimately contribute to improve public health measures against

vectorborne infections.

References

Ahmed MM., 2010. Clinical profile of Dengue fever infection in King Abdul Aziz University

Hospital Saudi Arabia. J Infect Dev. 4, 503-510.

Ciccozzi M., Callegaro A., lo Presti A., Cella E., Giovanetti M., Salpini R., Babakir-mina M.,

Farina C., Maggiolo F., Perno C.F., Ciotti M., 2013. When phylogenetic analysis complements the

epidemiological investigation: a case of HIV-2 infection, Italy. New Microbiol. 36(1), 93-6.

Drummond, A.J., Rambaut, A., Shapiro, B., Pybus, O.G., 2005. Bayesian coalescent inference of

past population dynamics from molecular sequences. Mol. Biol. Evol. 22, 1185–1192

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

14

Drummond, A.J., Rambaut, A., 2007. BEAST: bayesian evolutionary analysis by sampling trees.

BMC Evol. Biol. 7, 214.

EpiSouth., 2011. EpiSouthWeekly Epi Bulletin 155 (http://www.episouth.org).

Ghouth ASB, Amarasingh, A, Letson W., 2012. Dengue outbreak in Hadramouth, Yemen, 2010: an

epidemiological perspective. Am J Trop Med Hyg. 86, 1072-1076.

Hotez PJ, Savioli L, Fenwick A., 2012. Neglected tropical diseases of the Middle East and North

Africa: review of their prevalence, distribution, and opportunities for control. PLOS neglected

Tropical Diseases. 6, e1475.

Jimenez-Lucho VE, Fisher EJ, Saravolatz LD., 1984. Dengue with hemorrhagic manifestations: an

imported case from the Middle East. Am J Trop Med Hyg. 33, 650-653.

Kholedi AA, Balubaid O, Milaat W, Kabbash IA, Ibrahim A., 2012. Factors associated with the

spread of Dengue fever in Jeddah Governorate, Saudi Arabia. East Mediterr Health. 18, 15-23.

Madani TA, El-Tayeb ME, Al-Bar HMS, Azhar EI, Kao M, Alshoeb HO, Bamoosa AR., 2013.

Outbreak of viral hemorrhagic fever caused by Dengue virus type 3 in Al-Mukalla, Yemen. BMC

Infect Dis. 13, 136.

Nicoletti L, Ciccozzi M, Marchi A, Fiorentini C, Martucci P, D'Ancona F, Ciofi degli Atti M,

Pompa MG, Rezza G, Ciufolini MG., 2008. Chikungunya and Dengue Viruses in Travelers. Emerg

Infect Dis. Jan. 14(1), 177–178.

Ooi EE, Gubler DJ., 2011. Dengue and Dengue hemorrhagic fever (Ch. 75). In: Guerrant RL,

Walker DH, eds. Tropical infectious diseases: principles, pathogens and practice. 3rd

edition

Saunders Elsevier.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

15

Ravanini P, Huhtamo E, Hasu E, Rosa F, Costantino S, Crobu MG, Ilaria V, Nicosia AM, Garavelli

PL, Vapalahti O., 2010. Imported Dengue virus serotype 3, Yemen to Italy, 2010. Em Infect Dis.

17, 929-931.

Rezza G., El-Sawaf G., Faggioni G., Vescio F., Al Ameri R., De Santis R., Helaly G., Pomponi A.,

Metwally D., Fantini M., Qadi H., Ciccozzi M., Lista F., 2014. Co-circulation of Dengue and

Chikungunya Viruses, Al Hudaydah, Yemen, 2012. Emerg Infect Dis. 20(8), 1351-4.

World Health Organization., 2009. Dengue: guidelines for diagnosis, treatment, prevention and

control. New Ed. Geneva, Switzerland.

Zayed A, Awash AA, Esmal MA, Al-Mohamadi HA, Al-Salwai M, Al-Jasari A, Medhat

I, Morales-Betoulle ME, Mnzava A., 2012. Detection of Chikungunya virus in Aedes Aegypti

during the 2011 outbreak in Al Hodayda, Yemen. Acta Topica. 123, 62-66.

Zehender G, Ebranati E, Bernini F, Lo Presti A, Rezza G, Delogu M, Galli M, Ciccozzi M., 2011.

Phylogeography and epidemiological history of West Nile virus genotype 1a in Europe and the

Mediterranean basin. Infect Genet Evol. 11(3), 646-53.

Figure legends

Fig. 1. Maximum likelihood tree including 9 isolates of Dengue NS1 gene plus 49 reference

sequences. The isolate analyzed are in red. The tree was rooted by the midpoint rooting. The scale

bar at the bottom indicating 0.4 nucleotide substitutions per site. The • along the branches

represents significant statistical support for the clusters subtending that branch (p<0.001 in the zero-

branch-length test and bootstrap support ≥ 70%). Different lineages and subtypes are indicated by

bracket.

Fig. 2. Maximum likelihood phylogenetic analysis of 8 CHIKV partial E1 isolates from Yemen

and 111 reference sequences. The scale bar at the bottom indicates 0.02 nucleotide substitutions

per site. The • along the branch represents significant statistical support for the clade subtending

that branch (bootstrap ≥ 70%). The isolates from Yemen are indicated in bold.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

16

Fig. 3. Bayesian phylogenetic tree of 70 CHIKV partial E1 sequences . One • along the branches

represents significant statistical support for the clade subtending that branch (posterior probabilities

≥ 98%). Years are reported on the scale axis below. Isolates from Yemen are colored in red. The

cluster including the sequences Yemen is zoomed and his date is reported in the node. The Valine

or the Alanine in the aminoacid position 226 of the E1 sequences was indicated.

Fig.4. Bayesian phylogeographic tree of Dengue 2 cosmopolitan NS1 region sequences. The

significant statistical support for the clade or cluster subtending the branch (posterior probability ≥

98%) is indicated with a black circle along that branch. Geographic locations showed with different

colors in the tree are represent in the legend on the left. In the right corner there is the zoom of the

Yemen clade. The time of the most recent common ancestor, and the credibility interval based on

95% highest posterior density interval (95% HPD) is also reported.

Fig. S1. Likelihood mapping of the first (A), second (B) Dengue NS1 dataset. The dots inside the

triangles represents the posterior probabilities of the possible unrooted topologies for each quartet.

Numbers indicate the percentage of dots in the centre of the triangle corresponding to phylogenetic

noise (star-like trees).

Fig. S2. Likelihood mapping of CHIKV partial E1 sequences of the first (a) and second (b)

datasets. The dots inside the triangles represent the posterior probabilities of the possible unrooted

topologies for each quartet. Numbers indicate the percentage of dots in the centre of the triangle

corresponding to phylogenetic noise (star-like trees).

Figure(s)Click here to download high resolution image

Figure(s)Click here to download high resolution image

Figure(s)Click here to download high resolution image

Figure(s)Click here to download high resolution image

Supplementary MaterialClick here to download Supplementary Material: Fig. S1.tif

Supplementary MaterialClick here to download Supplementary Material: Fig. S2.tif