The Effect of Extensive Mosquito Migration on Dengue Control as Revealed by Phylogeny of Dengue Vector Aedes aegypti

Abstract

Dengue has become the most important arbo-viral disease in all tropical and subtropical regions of the world. Aedes aegypti, being the principal vector of the virus, vary in both epidemiological and behavioral characteristics, which could be finely measured through DNA sequence comparison at their population level. Such knowledge in the population differences can assist in implementation of efficient vector control strategies allowing to make estimates of the gene flow and adaptive genomic changes, which are important predictors of spread of Wolbachia infection or insecticide resistance. As such, this study was undertaken to investigate the phylogenetic relationships of Ae. aegypti from Galle and Colombo, Sri Lanka, based on the ribosomal protein region which spans between two exons, in order to understand the geographical distribution of genetically distinct mosquito clades and its impact on mosquito control measures. A 320bp DNA region spanning from 681-930 bp, corresponding to the ribosomal protein, was sequenced in 62 Ae. aegypti larvae collected from Galle (N=30) and Colombo (N=32), Sri Lanka. The sequences were aligned using ClustalW and the haplotypes were determined with DnaSP 5.10. Phylogenetic relationships among haplotypes were constructed using the maximum likelihood method under Tamura 3 parameter model in MEGA 7.0.14 including three previously reported sequences of Australian (N=2) and Brazilian (N=1) Ae. aegypti. The bootstrap support was calculated using 1000 replicates and the tree was rooted using Aedes notoscriptus (GenBank accession No. KJ194101). Among all sequences, nineteen different haplotypes were found among which 5 haplotypes were shared between 80% of mosquitoes in the two populations. Seven haplotypes were unique to each of the population. Phylogenetic tree revealed two basal clades and a single derived clade. All observed haplotypes of the two Ae. aegypti populations were distributed in all the three clades, indicating a lack of genetic differentiation between populations. The Brazilian Ae. aegypti haplotype and one of the Australian haplotypes were grouped together with the Sri Lankan basal haplotype in the same basal clade, whereas the other Australian haplotype was found in the derived clade. Phylogram showed that Galle and Colombo Ae. aegypti populations are highly related to each other despite the large geographic distance (129 Km) indicating a substantial genetic similarity between them. This may have probably arisen from passive migration assisted by human travelling and trade through both land and water as the two areas are bordered by the sea. In addition, studied Sri Lankan mosquito populations were closely related to Australian and Brazilian samples. Probably this might have caused by shipping industry between the three countries as all of them are fully or partially enclosed by sea. For example, illegal fishing boats migrating to Australia by sea is perhaps a good mean of transportation of all life stages of mosquitoes from Sri Lanka. These findings indicate that extensive mosquito migrations occur between populations not only within the country, but also among other countries in the world which might be a main barrier to the successful vector control measures. M. D. Nirmani is with the

Full text

Abstract— Dengue has become the most important arbo-viral
disease in all tropical and subtropical regions of the world. Aedes
aegypti, being the principal vector of the virus, vary in both
epidemiological and behavioral characteristics, which could be finely
measured through DNA sequence comparison at their population level.
Such knowledge in the population differences can assist in
implementation of efficient vector control strategies allowing to make
estimates of the gene flow and adaptive genomic changes, which are
important predictors of spread of Wolbachia infection or insecticide
resistance. As such, this study was undertaken to investigate the
phylogenetic relationships of Ae. aegypti from Galle and Colombo, Sri
Lanka, based on the ribosomal protein region which spans between
two exons, in order to understand the geographical distribution of
genetically distinct mosquito clades and its impact on mosquito control
measures. A 320bp DNA region spanning from 681-930 bp,
corresponding to the ribosomal protein, was sequenced in 62 Ae.
aegypti larvae collected from Galle (N=30) and Colombo (N=32), Sri
Lanka. The sequences were aligned using ClustalW and the haplotypes
were determined with DnaSP 5.10. Phylogenetic relationships among
haplotypes were constructed using the maximum likelihood method
under Tamura 3 parameter model in MEGA 7.0.14 including three
previously reported sequences of Australian (N=2) and Brazilian
(N=1) Ae. aegypti. The bootstrap support was calculated using 1000
replicates and the tree was rooted using Aedes notoscriptus (GenBank
accession No. KJ194101). Among all sequences, nineteen different
haplotypes were found among which 5 haplotypes were shared
between 80% of mosquitoes in the two populations. Seven haplotypes
were unique to each of the population. Phylogenetic tree revealed two
basal clades and a single derived clade. All observed haplotypes of the
two Ae. aegypti populations were distributed in all the three clades,
indicating a lack of genetic differentiation between populations. The
Brazilian Ae. aegypti haplotype and one of the Australian haplotypes
were grouped together with the Sri Lankan basal haplotype in the same
basal clade, whereas the other Australian haplotype was found in the
derived clade. Phylogram showed that Galle and Colombo Ae. aegypti
populations are highly related to each other despite the large
geographic distance (129 Km) indicating a substantial genetic
similarity between them. This may have probably arisen from passive
migration assisted by human travelling and trade through both land and
water as the two areas are bordered by the sea. In addition, studied Sri
Lankan mosquito populations were closely related to Australian and
Brazilian samples. Probably this might have caused by shipping
industry between the three countries as all of them are fully or partially
enclosed by sea. For example, illegal fishing boats migrating to
Australia by sea is perhaps a good mean of transportation of all life
stages of mosquitoes from Sri Lanka. These findings indicate that
extensive mosquito migrations occur between populations not only
within the country, but also among other countries in the world which
might be a main barrier to the successful vector control measures.
M. D. Nirmani is with the Department of Zoology and Environment Science,
Faculty of Science, University of Colombo, No. 94, Cumarathunga Munidasa
Mawatha, Colombo 03, Sri Lanka (e-mail: dhanushanirmani@gmail.com).
K. L. N. Perera is with the Genetech Molecular Diagnostics and School of
Gene Technology, No. 54, Kithulwaththe Road, Colombo 08, Sri Lanka (e-
mail: nanadikasp@hotmail.com).
Keywords—Aedes aegypti, dengue control, extensive mosquito
migration, haplotypes, phylogeny, ribosomal protein
G. H. Galhena is with the Department of Zoology and Environment Science,
Faculty of Science, University of Colombo, No. 94, Cumarathunga Munidasa
Mawatha, Colombo 03, Sri Lanka (phone: +94112508372; fax: +94112586868;
e-mail: gayanihg@gmail.com).
M. D. Nirmani, K. L. N. Perera, G. H. Galhena
The Effect of Extensive Mosquito Migration on Dengue Control as Revealed by Phylogeny
of Dengue Vector Aedes aegypti