Scott C. Weaver Institute for Human Infections and Immunity

and Department of Microbiology and Immunology,

University of Texas Medical Branch, Galveston

Chikungunya-Vector Relationships and Prospects for Control in

Americas and Beyond

Aedes  spp.  

Ae.  aegyp+  aegyp+  Ae.  albopictus  

Sub-Saharan Africa

TSETSARKIN, K. A., CHEN, R., SHERMAN, M. B. & WEAVER, S. C. 2011. Curr Opin Virol, 1, 310-317.

No known barriers to initial

host range changes

Kraemer, M.U., et al., 2015. The global distribution of the arbovirus vectors Aedes aegypti and Ae. albopictus. Elife 4.

Precipitation and vegetation indices made up the remainder of predictors. Urban land cover madevery little contribution to either model (Table 2). Model evaluation statistics under cross-validationwere high (AUC: 0.87 and 0.9 respectively) for both model ensembles, indicating high predictiveperformance of the model. Effect plots for each covariate are shown in Figure 1—figure supplement 2.Maps of uncertainty associated with these predictions are presented in Figure 1—figure supplement 3.

DiscussionBy combining the most comprehensive dataset of occurrence records with an advanced modellingapproach and a bespoke set of environmental and land-cover correlates, we have producedcontemporary high-resolution probability of occurrence maps for Ae. aegypti and Ae. albopictus, twoof the most important disease vectors globally. Dengue and chikungunya, pathogens transmitted bythese vectors and rapidly expanding in their distributions, are increasingly prominent in public healthagendas and pose significant health threats to humans (Staples et al., 2009; Gardner et al., 2012;Bhatt et al., 2013; Weaver and Lecuit, 2015). In common with previous work to map the globaldistributions of the dominant vectors of malaria (Sinka et al., 2010a, 2010b, 2011), the maps willimprove efforts to understand the spatial epidemiology of associated arboviruses, and to predict howthese could change in the future. Specifically, these maps may be used to prioritize surveillance forthese vector species and the diseases caused by the viruses they transmit in areas where disease andentomological reporting remains poor. For example, in parts of Asia and Africa where there isa mismatch between predicted environmental suitability and reported occurrences, these maps couldbe used to determine whether the vector has yet to fill its niche or if it is present but has not beenreported due to limited entomological surveillance. They may also be used to identify areas wherethe species could persist but has yet to be reported, in order to proactively prevent vectorestablishment.

The relative contributions of each of the environmental covariates to the global models concur withour theoretical and experimental understanding of each species’ biology. Both species’ distributionsare highly dependent on the limiting factor temperature places on survival of the adult mosquitoesand on the gonotrophic cycle (Brady et al., 2013) (Table 2). The inclusion of a bespoke temperaturesuitability index (Brady et al., 2014), both in defining the pseudo-absences and as a covariate,allowed us to capture both geographic and temporal variations in the species-specific effects oftemperature in a single variable, leading to improved predictive skill of the models. As both

Figure 2. Global map of the predicted distribution of Ae. albopictus. The map depicts the probability of occurrence (from 0 blue to 1 red) at a spatial

resolution of 5 km × 5 km.

DOI: 10.7554/eLife.08347.009

Kraemer et al. eLife 2015;4:e08347. DOI: 10.7554/eLife.08347 6 of 18

Research article Ecology | Epidemiology and global health

A. albopictus

Originated in Asia, spread to the Americas, Africa and Europe beginning in 1985

A. aegypti

In Europe, the predicted potential distribution of Ae. albopictus contains most of the knownoccurrence points, but suitability is also predicted in Portugal and the west of Spain, and in much ofsouth-eastern Europe and the Balkans, where the species has yet to be reported. Similarly, in ChinaAe. albopictus has yet to be reported from much of the area predicted to be environmentally suitable.By contrast, in the United States the species has been reported from almost all of the predictedsuitable areas, with the exception of a small band of predicted suitability on the western slope of theSierra Nevada. Due to the relatively sparse reporting from Africa it remains uncertain whether areaspredicted to be highly suitable are already infested or have yet to be colonized by the species.Ae. albopictus for example has only been reported from some West African countries (Nigeria,Cameroon, Gabon, the Central African Republic, Congo, Cote d’Ivoire) and Madagascar, and SouthAfrica (as well as some islands in the Indian Ocean). The distribution of Ae. aegypti in Africa seemsto be much wider, with reports of species occurrence in over 30 countries.

For both species, the most important predictor was temperature. Temperature suitability indiceshad high relative influence statistics for both species; this variable was selected in approximatelyhalf of regression tree decisions for Ae. aegypti (54.9%, CI = 53.7–56%) and Ae. albopictus (44.3%,CI = 42.7–45.6%). The full definition of a relative influence statistic is given in the ‘Materials andmethods’ section under the heading Predictive performance and relative influence of covariates.

Figure 1. Global map of the predicted distribution of Ae. aegypti. The map depicts the probability of occurrence (from 0 blue to 1 red) at a spatial

resolution of 5 km × 5 km.

DOI: 10.7554/eLife.08347.004

The following figure supplements are available for figure 1:

Figure supplement 1. Effect plots of covariates used in this study showing the marginal effect of each covariate on probability of presence for Ae. aegypti

(1) and Ae. albopictus (2): enhanced vegetation index (EVI) annual mean (A); Enhanced vegetation index—range (B); annual monthly maximum

precipitation (C); annual monthly minimum precipitation (D); temperature suitability (E); urban areas (F); peri-urban areas (G).

DOI: 10.7554/eLife.08347.005

Figure supplement 2. Set of covariate layers used to predict the ecological niche of Ae. aegypti and Ae. albopictus described in detail in the ‘Materials

and methods’ section; (A) enhanced vegetation index (EVI) annual mean, (B) EVI annual range, (C) annual monthly maximum precipitation, (D) annual

monthly minimum precipitation, (E) temperature suitability for Ae. albopictus, (F) temperature suitability for Ae. aegypti, (G) rural, peri-urban and urban

classification layer.

DOI: 10.7554/eLife.08347.006

Figure supplement 3. Visualization of pixel level uncertainty calculated using the upper and lower bounds of the 95% confidence intervals associated with

the prediction maps for Ae. aegypti (A) and Ae. albopictus (B).

DOI: 10.7554/eLife.08347.007

Figure supplement 4. The distribution of the occurrence database for Ae. aegypti (A) and Ae. albopictus (B) plotted on the underlying prediction surface.

DOI: 10.7554/eLife.08347.008

Kraemer et al. eLife 2015;4:e08347. DOI: 10.7554/eLife.08347 5 of 18

Research article Ecology | Epidemiology and global health

Originated in sub-Saharan Africa, spread throughout the tropics centuries ago

Urban Chikungunya Virus Vectors

CHIKV Endemic/Epidemic Vectors

Aedes aegypti aegypti •  Tropical and subtropical •  Feeds almost exclusively on humans •  Takes multiple bloodmeals within a

gonotrophic cycle •  Exploits artificial water containers near

houses as larval habitats •  Adult females found mostly inside houses •  Feeds during the daytime •  Moderately susceptible to CHIKV

Aedes albopictus •  Tropics and temperate regions •  Feeds opportunistically •  Usually takes a single bloodmeal within a

gonotrophic cycle •  Uses artificial and natural larval habitats •  Varied levels of anthrophily and endophily •  Feeds during the daytime •  Moderately to highly susceptible to CHIKV

behavior susceptibility

CHIKV Evolution

!"

!#

$%

&'()(*&+(',-./

!

&012

" # $%% $&% '(%)*+,-./012

3 34&5.'(6.7!"8&!#8&!9

'()(*&+(',-./

6,)7(*&:$%;

$%<012&=(>?@/.

"A&0,*(,'&A.B7(?-&%/CD('.&

A(7?,B6.&D/C=&?5.&6.B?./&:E;

-0.5

0

0.5

1

1.5

2

0 50 100 150 200 250 300 350

""9

"F9

'()(*&+(',-./$%<012=(>?@/.

4

!9

!"

!#

$%

&'()(*&+(',-./

!

&012

" # $%% $&% '(%)*+,-./012

3 34&5.'(6.7!"8&!#8&!9

'()(*&+(',-./

6,)7(*&:$%;

$%<012&=(>?@/.

"A&0,*(,'&A.B7(?-&%/CD('.&

A(7?,B6.&D/C=&?5.&6.B?./&:E;

-0.5

0

0.5

1

1.5

2

0 50 100 150 200 250 300 350

""9

"F9

'()(*&+(',-./$%<012=(>?@/.

4

!9

lipid bilayers glycoproteins E1 & E2

capsid

RNA

Fig.!3!

nsP1 nsP2 nsP3 nsP4 C E2 E1 Poly[A]-3’ 5’ cap

26S

(-) strand RNA synthesis,

RNA capping

Helicase, proteinase

RNA synthesis

RNA-dependent RNA

polymerase Envelope

glycoproteins

6K

E3

Capsid

Alphavirus Genome and Structure

Zhang, R., et al., 2011. EMBO J 30, 3854-3863.

Indian Ocean epidemics 2005-2011

Asia 2005-2009

Convergent Evolution of E1-A226V Substitution

E1-A226V

Tsetsarkin, K. A., Chen, R., Sherman, M., and Weaver, S. C. (2011). Curr Opin Virol 1, 310-317 .

Enzootic strains East, Central, South Africa

Asian epidemics 1958-2006

Enzootic strains West Africa

226V 226A

Effect of E1-A226V Mutation on Aedes albopictus Infectivity

OID50 expressed as Log10TCID50/ml

Schuffenecker I., et al. PLoS Med. 2006;3(7):e263. Vazeille, M., et al., 2007. PLoS ONE 2, e1168. Tsetsarkin, K.A., et al., 2007. PLoS Pathog 3, e201.

A. albopictus (E1-226V)

Human Infection Profile for CHIKV Ti

ter (

Log 1

0 inf

ectio

us u

nits

/ml s

erum

)

Days after infection 1 2 3 4 5 6 7 8 9 10

Viremia IgM IgG

2

4

6

8

10 Transmission window

Antibody titer

A. aegypti, A. albopictus (E1-226A)

A. albopictus-adaptive mutations

p=0.005

4.86

41.5

4.75

43.5

E2 E1 substitutions enhance initial CHIKV infection of A. albopictus midgut cells

The E2-L201Q fitness increase is ca. 10-fold weaker than E1-A226V and was selected by A. albopictus only after ca. 4 years of circulation in India

Neither E1-A226V nor E2-L201Q affects infection of A. aegypti

Tsetsarkin, K.A., Weaver, S.C., 2011. PLoS Pathog 7, e1002412

E3; E2; E1 proteins Natural, second-step adaptive substitutions of glutamine in acid-sensitive 210-252 E2 region

A. albopictus-adaptive mutations in the acid sensitive region of E2

Hypothesis: Gln or Glu substitutions in the E2 acid-sensitive region enhance E1-226V by regulating low pH-induced E2-E1 heterodimer disassociation for efficient CHIKV endosomal entry in A. albopictus midgut cells. Support: E2. substitutions are not in the receptor-binding domain, and artificial Gln substitutions in the acid sensitive region also enhance A. albopictus infectivity.

Tsetsarkin KA, et al. Nat Commun. 2014;5:4084. Epub 2014/06/17.

Indian Ocean epidemics 2005-2011

Asia 2005-2009

Convergent Evolution of E1-A226V Substitution

E1-226V

Tsetsarkin, K. A., Chen, R., Sherman, M., and Weaver, S. C. (2011). Curr Opin Virol 1, 310-317 .

Enzootic strains East, Central, South Africa

Asian epidemics 1958-2006

Enzootic strains West Africa

X

Conundrum: Why has A226V mutation not been selected in Southeast Asia despite endemic circulation in native A. albopictus territory for ≥57 years? Explanation: E1-226V does not confer a fitness advantage in Asian CHIKV strains due to epistatic interactions with E1-98T. This amino acid, only found in Asian genotype strains, prevents probably resulted from a founder effect when this genotype was introduced from eastern Africa in the late 19th or early 20th Century. Prediction: Most CHIKV strains now circulating in the Americas, which belong to the Asian genotype, will be more efficiently transmitted by A. aegypti

Prospects for the Control of Chikungunya Spread

1.  Reduction of contact between people and mosquito vectors

2.  Reduction of human amplification using antivirals

3.  Control of transmission by reducing populations of mosquito vectors

4.  Prevention of human infection using vaccines

Failure of Traditional Mosquito Control to Control Dengue

•  Adult female A. aegypti remain inside houses, limiting insecticide penetration from traditional applications

•  Larval A. aegypti are found in low density in artificial containers around homes, requiring entry into individual properties for inspection/control

•  A. aegypti in many parts of the world are developing resistance to insecticides

•  Tetracycline repressible activator variant (tTAV): acts as a switch to control the activity of essential mosquito genes

•  tTAV works only in insect cells; the non toxic protein ties up the cell’s machinery so it’s other genes aren’t expressed and the insect dies in the larval stages

•  Tetracycline binds to tTAV and disables it, allowing it to be added to laboratory larval water so that mosquitoes survive to the adult stage

•  Because the tTAV gene is dominant, offspring of matings between released transgenic and wild mosquitoes die in the wild without tetracycline in their larval habitats

Novel Approach for Vector Control: Release of Insects with Dominant Lethality (RIDL)

•  Safety advantage: transgene is “suicidal” •  Potential limitations: Sustained release of mosquitos is required,

may be too costly for resource-limited regions endemic to CHIKV

Implementation of RIDL: release of transgenic

male mosquitoes

Field trials: Cayman Islands, Malaysia,

Panama and Brazil

Wolbachia: Bacterial symbionts of many insects that can spread through populations through cytoplasmic incompatibility

Female Male

•  When adapted to and introduced into Aedes aegypti, Wolbachia reduce their lifespan

•  Wolbachia also reduce CHIKV (and dengue virus) replication and transmission

Field Trials: Australia (completed), Indonesia, Viet Nam, Brazil, Colombia

Potential limitations: Limited dispersal of A. aegypti will necessitate widespread release of female mosquitoes. Can CHIKV (and dengue virus) evolve resistance to the Wolbachia suppression?

CHIK Vaccine Development The good news: •  Single CHIKV serotype,

no evidence of reinfection or immune enhancement

•  Well established correlates of protection for alphaviruses (neutralizing antibodies)

•  Ease of genetic manipulation for rational attenuation

•  Good nonhuman primate models for human disease

The bad news: •  Immunocompetent

rodents are poor models for human disease

•  Unpredictable market after epidemics subside and CHIKV is rarely diagnosed

•  Unpredictable incidence of disease for clinical efficacy trials

•  FDA animal rule has not yet produced a licensed vaccine

Chikungunya Vaccine Development

Bharat Biotech; VLP vaccine

Takeda and UTMB: Recombinant live attenuated CHIK/

IRES

ArboVax; Recombinant LAIV

Inovio; DNA vaccine

Themis; Measles-based Recombinant

Indian Immunologicals; Walter Reed strain

US Army, 181/clone25 live attenuated vaccine

based on 2 point attenuating mutations

Indian Immunologicals; Inactivated strain 181/

clone 25

Preclinical Phase I Phase II Phase III

NIAID; VLP vaccine

manufactured in CHO cells

*Merck option ?

Yale, Profectus, UTMB: VSV-vectored

live-attenuated

Summary •  CHIKV has emerged repeatedly, probably for centuries, from its enzootic cycle in

Africa into an urban human-mosquito cycle in Asia, Europe and the Americas.

•  The 2005-2015 epidemic has been extensive, in part due to the sequential, convergent and step-wise adaptation of 5 natural A. albopictus-adaptive envelope glycoprotein gene mutations, with little or no effect on infection of A. aegypti or murine models of human infection.

•  An epistatic interaction between E1 residues 98 and 226 has prevented the adaptation of CHIKV for A. albopictus transmission in the Asian lineage, despite the E1-98 residues having no detectable phenotype. This epistasis should limit the ability of CHIKV strains in the Americas to adapt for A. albopictus transmission. An ECSA strain introduced into Brazil in 2014 from Angola also has an epistatic constraint that may prevent optimal adaptation to this new CHIKV vector.

•  Traditional methods of vector control are unlikely to prevent the spread of CHIKV. However, new approaches to vector reduction and interference with vector transmission are more promising

•  Several promising vaccines for CHIKV have been developed but financial and regulatory hurdles represent major obstacles to their licensure

Acknowledgements

Funding: NIH-NIAID R01-AI069145, R01-AI071192, R01-AI48807

Konstantin Tsetsarkin Rubing Chen Ruimei Yun

Indiana University Tuli Mukhopadhyay

Institut Pasteur Amadou Sall Mawlouth Diallo

Johns Hopkins University Derek Cummings Ben Althouse

NM State University Kathy Hanley